Search

We give systematic ways of defining monotone quantum relative entropies and (multi-variate) quantum Rényi divergences starting from a set of monotone quantum relative entropies. Interestingly, despite its central importance in information theory, only two additive and monotone quantum extensions of the classical relative entropy have been known so far, the Umegaki and the Belavkin-Staszewski relative entropies, which are the minimal and the maximal ones, respectively, with these properties. Using the Kubo-Ando weighted geometric means, we give a general procedure to construct monotone and additive quantum relative entropies from a given one with the same properties; in particular, when starting from the Umegaki relative entropy, this gives a new one-parameter family of monotone (even under positive trace-preserving (PTP) maps) and additive quantum relative entropies interpolating between the Umegaki and the Belavkin-Staszewski ones on full-rank states. In a different direction, we use a generalization of a classical variational formula to define multi-variate quantum Rényi quantities corresponding to any finite set of quantum relative entropies (D q x ) x ∈ X and real weights (P (x)) x ∈ X summing to 1, as Q P b , q ((ϱ x) x ∈ X) : = sup τ ≥ 0 ⁡ { Tr τ − ∑ x P (x) D q x (τ ‖ ϱ x) }. We analyze in detail the properties of the resulting quantity inherited from the generating set of quantum relative entropies; in particular, we show that monotone quantum relative entropies define monotone Rényi quantities whenever P is a probability measure. With the proper normalization, the negative logarithm of the above quantity gives a quantum extension of the classical Rényi α -divergence in the 2-variable case (X = { 0 , 1 } , P (0) = α). We show that if both D q 0 and D q 1 are lower semi-continuous, monotone, and additive quantum relative entropies, and at least one of them is strictly larger than the Umegaki relative entropy then the resulting barycentric Rényi divergences are strictly between the log-Euclidean and the maximal Rényi divergences, and hence they are different from any previously studied quantum Rényi divergence. [ABSTRACT FROM AUTHOR]

We study a setting where a receiver must design a questionnaire to recover a sequence of symbols known to a strategic sender, whose utility may not be incentive compatible. We allow the receiver the possibility of selecting the alternatives presented in the questionnaire, and thereby linking decisions across the components of the sequence. We show that, despite the strategic sender and the noise in the channel, the receiver can recover exponentially many sequences, but also that exponentially many sequences are unrecoverable even by the best strategy. We define the growth rate of the number of recovered sequences as the information extraction capacity. A generalization of the Shannon capacity, it characterizes the optimal amount of communication resources required while communicating with a strategic sender. We derive bounds leading to an exact evaluation of the information extraction capacity in many cases. Our results form the building blocks of a novel, non-cooperative regime of communication involving a strategic sender. • An uninformed receiver aims to recover information from an informed strategic sender • No assumption of incentive compatibility or external incentives • Information-theoretic analysis shows a fundamental limit on recovered information • Real-world setting of screening of travellers using questionnaires can be studied • Novel theory combining Shannon's information theory and Myerson's mechanism design [ABSTRACT FROM AUTHOR]

In this paper, we study algorithmic questions concerning products of matrices and their consequences for recognition algorithms for polyhedra. The 1- product of matrices S 1 ∈ R m 1 × n 1 , S 2 ∈ R m 2 × n 2 is a matrix in R (m 1 + m 2) × (n 1 n 2) whose columns are the concatenation of each column of S 1 with each column of S 2. The k - product generalizes the 1-product, by taking as input two matrices S 1 , S 2 together with k − 1 special rows of each of those matrices, and outputting a certain composition of S 1 , S 2 . Our first result is a polynomial time algorithm for the following problem: given a matrix S , is S a k -product of some matrices, up to permutation of rows and columns? Our algorithm is based on minimizing a symmetric submodular function that expresses mutual information from information theory. Our study is motivated by a close link between the 1-product of matrices and the Cartesian product of polytopes, and more generally between the k -product of matrices and the glued product of polytopes. These connections rely on the concept of a slack matrix, which gives an algebraic representation of classes of affinely equivalent polytopes. The slack matrix recognition problem is the problem of determining whether a given matrix is a slack matrix. This is an intriguing problem whose complexity is unknown. Our algorithm reduces the problem to instances which cannot be expressed as k -products of smaller matrices. In the second part of the paper, we give a combinatorial interpretation of k -products for two well-known classes of polytopes: 2-level matroid base polytopes and stable set polytopes of perfect graphs. We also show that the slack matrix recognition problem is polynomial-time solvable for such polytopes. Those two classes are special cases of 2-level polytopes, for which we conjecture that the slack matrix recognition problem is polynomial-time solvable. [ABSTRACT FROM AUTHOR]

Diverse explanations or theories of consciousness are arrayed on a roughly physicalist-to-nonphysicalist landscape of essences and mechanisms. Categories: Materialism Theories (philosophical, neurobiological, electromagnetic field, computational and informational, homeostatic and affective, embodied and enactive, relational, representational, language, phylogenetic evolution); Non-Reductive Physicalism; Quantum Theories; Integrated Information Theory; Panpsychisms; Monisms; Dualisms; Idealisms; Anomalous and Altered States Theories; Challenge Theories. There are many subcategories, especially for Materialism Theories. Each explanation is self-described by its adherents, critique is minimal and only for clarification, and there is no attempt to adjudicate among theories. The implications of consciousness explanations or theories are assessed with respect to four questions: meaning/purpose/value (if any); AI consciousness; virtual immortality; and survival beyond death. A Landscape of Consciousness, I suggest, offers perspective. [ABSTRACT FROM AUTHOR]

Schizophrenia is a polygenic complex disease with a heritability as high as 80 %, yet the mechanism of polygenic interaction in its pathogenesis remains unclear. Studying the interaction and regulation of schizophrenia susceptibility genes is crucial for unraveling the pathogenesis of schizophrenia and developing antipsychotic drugs. Therefore, we developed a bioinformatics method named GRACI (Gene Regulation Analysis based on Causal Inference) based on the principles of information theory, a causal inference model, and high order chromatin 3D conformation. GRACI captures the interaction and regulatory relationships between schizophrenia susceptibility genes by analyzing genotyping data. Two datasets, comprising 1459 and 2065 samples respectively, were analyzed, and the gene networks from both datasets were constructed. GRACI showcased superior accuracy when compared to widely adopted methods for detecting gene-gene interactions and intergenic regulation. This alignment was further substantiated by its correlation with chromatin high-order conformation patterns. Using GRACI, we identified three potential genes— KCNN3 , KCNH1 , and KCND3 —that are directly associated with schizophrenia pathogenesis. Furthermore, the results of GRACI on the standalone dataset illustrated the method's applicability to other complex diseases. GRACI download: https://github.com/liuliangjie19/GRACI [ABSTRACT FROM AUTHOR]

Historically, economic models of memory have neglected a pivotal aspect: decay, the established notion that memories deteriorate over time, resulting in a loss of fidelity. In this paper we show that a framework of rational memory with decay can produce the recency effect and other intuitively appealing memory phenomena. We apply the framework to both finite time and dynamic programming settings, deriving a wide range of results with a focus on comparative statics and asymptotic behaviors. Many of our results can also be applied to multi-dimensional rational inattention settings where individuals have to split their attention across several information sources. [ABSTRACT FROM AUTHOR]

• Trait of aggression and aggressive driving behaviors are positively correlated. • SIP deficits in daily and driving contexts are positively and strongly correlated. • Individuals attribute significantly more hostility in driving than in daily contexts. • Self-efficacy as the most important SIP to act aggressively. Questions such as "Are people always aggressive?" or "Is it only inside the car?" have been studied in previous research. However, the focus is mostly on anger and anger expression and not on an interpersonal perspective. The current study aims to explore this gap in the research by analyzing aggressive driving behavior through the application of Social Information Processing Theory (SIP). For that purpose, the trait of aggression was compared with aggressive driving behavior and SIP patterns on and off the road. To achieve these goals, 562 Portuguese drivers completed a self-reported survey concerning trait aggression, aggressive driving behavior, and social information processing in various contexts. The results indicated that deficits in SIP positively correlate with aggressive behavior in daily and driving contexts. Secondly, the trait of aggression is positively correlated with aggressive driving behavior. Finally, SIP deficits in the daily context were also positively correlated with SIP deficits in the driving context. Nevertheless, significant differences were found between SIP variables in the daily context and the driving context. The results, implications, and limitations of the study are discussed. [ABSTRACT FROM AUTHOR]

Collisions involving space objects pose significant threats to space sustainability. Within these collisions, catastrophic collisions which generate a larger amount of debris are particularly harmful. The quantity and characteristics of fragments generated from a collision are usually modeled using the NASA standard breakup model (SBM). However, on several occasions, the NASA SBM has incorrectly estimated the effects of collisions, including the Yunhai 1-02 breakup event. Using outer probability measure theory and information theory, we develop a correction methodology to estimate the Yunhai 1-02 event based on NASA SBM. This methodology successfully passes a statistical test using historical data. After correction, the characteristic diameter and relative velocity distributions of fragments generated from this event exhibit greater similarity with observations. The work in this paper facilitates the analysis of the properties of the debris cloud resulting from collisions with limited information, improving the predictions of future collisions. [Display omitted] • We condense and analyze the historical collision data. • We obtain a corrected parameter to adjust the SBM result, and this adjusted result passes the statistical test. • We improve the precision of characteristic diameter and relative velocity of fragments generated from Yunhai event. • This work is important to estimate the effect of collisions with limited information. [ABSTRACT FROM AUTHOR]

This study presents a methodological framework for optimal placement of pressure sensors in Water Distribution Networks (WDNs) for leakage monitoring under uncertainty. Monte Carlo simulation is utilized to simulate leakages of different magnitudes at various nodes in the WDN taking into consideration background noise and minimum resolution of pressure sensors. A novel sensor preselection algorithm based on community detection and maximum entropy computation to reduce the search space of the pressure Sensor Placement Problem (SPP) is presented. The pressure SPP is formulated as a multi-objective optimization problem that seeks to maximize Joint Entropy, Coverage, and minimize Total Correlation. NSGA-II is used to solve the SPP and the solutions in the optimal Pareto front are ranked using a hybrid Entropy TOPSIS to eliminate potential bias and subjective human judgement in optimal sensor configuration implementation. The sensor preselection algorithm achieved a 67% reduction in the search space (possible sensor positions) of the case study, C-TOWN WDN, with only 2.78% reduction in coverage. The result of the pressure SPP indicates only 21 pressure sensors are needed to cover 95.45% of the WDN under study. Finally, the overall performance of the proposed methodological framework is presented and compared with other related works. • A multiobjective optimal pressure sensor placement problem is formulated. • Sensor preselection algorithm based on Community Detection & Entropy is introduced. • Total correlation as a measure of redundancy ensures sparsity between sensors. • Entropy-TOPSIS eliminates bias in implemetation of optimal sensor configuration. • Limited number of pressure sensors can be placed optimally in distribution systems. [ABSTRACT FROM AUTHOR]

Information is not a monolithic entity, but can be decomposed into synergistic, unique, and redundant components. Relative predominance of synergy and redundancy in the human brain follows a unimodal–transmodal organisation and reflects underlying structure, neurobiology, and dynamics. Brain regions navigate trade-offs between these components to combine the flexibility of synergy for higher cognition and the robustness of redundancy for key sensory and motor functions. Redundancy appears stable across primate evolution, whereas synergy is selectively increased in humans and especially in human-accelerated regions. Computational studies offer new insights into the causal relationship between synergy, redundancy, and cognitive capabilities. To explain how the brain orchestrates information-processing for cognition, we must understand information itself. Importantly, information is not a monolithic entity. Information decomposition techniques provide a way to split information into its constituent elements: unique, redundant, and synergistic information. We review how disentangling synergistic and redundant interactions is redefining our understanding of integrative brain function and its neural organisation. To explain how the brain navigates the trade-offs between redundancy and synergy, we review converging evidence integrating the structural, molecular, and functional underpinnings of synergy and redundancy; their roles in cognition and computation; and how they might arise over evolution and development. Overall, disentangling synergistic and redundant information provides a guiding principle for understanding the informational architecture of the brain and cognition. [ABSTRACT FROM AUTHOR]

Recent theories and empirical evidence demonstrate that early sensory systems should encode abstract representations beyond 'realism'. This flexibility significantly influences our preferences and supposedly irrational behavior, as well as our emotional categorization, aesthetics, pleasure, and risk attitudes. However, these views leave it unclear at which level of the neural processing stream, and at which temporal scale, organisms should adopt efficient adaptations of the sensory information. Survival and optimal behavior also hinge on a 'sixth sense' – sensing the internal state of the body. Emerging evidence suggests that early sensory encoding of these signals efficiently helps to extract arousal, emotions, and behavioral rules. Is the role of our sensory systems to represent the physical world as accurately as possible? If so, are our preferences and emotions, often deemed irrational, decoupled from these 'ground-truth' sensory experiences? We show why the answer to both questions is 'no'. Brain function is metabolically costly, and the brain loses some fraction of the information that it encodes and transmits. Therefore, if brains maximize objective functions that increase the fitness of their species, they should adapt to the objective-maximizing rules of the environment at the earliest stages of sensory processing. Consequently, observed 'irrationalities', preferences, and emotions stem from the necessity for our early sensory systems to adapt and process information while considering the metabolic costs and internal states of the organism. [ABSTRACT FROM AUTHOR]

In a recent paper, Aceves and Evans computed information and semantic density measures for hundreds of languages, and showed that these measures predict the pace and breadth of ideas in communication. Here, we summarize their key findings and situate them in a broader debate about the adaptive nature of language. [ABSTRACT FROM AUTHOR]

Artificial Intelligence (AI), as an academic discipline, is traceable to the mid-1950s but it is currently exploding in applications with successes and concerns. AI can be defined as intelligence demonstrated by computers, with intelligence difficult to define but it must include concepts of ability to learn, reason, and generalize from a vast amount of information and, we propose, to infer meaning. The type of AI known as general AI, has strong, but unrealized potential both for assessing and also for solving major problems with the scientific theory of Darwinian evolution, including its modern variants and for origin of life studies. Specifically, AI should be applied first to evaluate the strengths and weaknesses of the assumptions and empirical information underpinning theories of the origin of life and probability of its evolution. AI should then be applied to assess the scientific validity of the theory of how abundant life came to be on earth. [ABSTRACT FROM AUTHOR]

The statistical supervised learning framework assumes an input–output set with a joint probability distribution that is reliably represented by the training dataset. The learning system is then required to output a prediction rule learned from the training dataset's input–output pairs. In this work, we investigate the relationship between the sample complexity, the empirical risk and the generalization error based on the asymptotic equipartition property (AEP) (Shannon, 1948). We provide theoretical guarantees for reliable learning under the information-theoretic AEP, with respect to the generalization error and the sample size in different settings. [ABSTRACT FROM AUTHOR]

Multi-view representation learning aims to capture comprehensive information from multiple views of a shared context. Recent works intuitively apply contrastive learning to different views in a pairwise manner, which is still scalable: view-specific noise is not filtered in learning view-shared representations; the fake negative pairs, where the negative terms are actually within the same class as the positive, and the real negative pairs are coequally treated; evenly measuring the similarities between terms might interfere with optimization. Importantly, few works study the theoretical framework of generalized self-supervised multi-view learning, especially for more than two views. To this end, we rethink the existing multi-view learning paradigm from the perspective of information theory and then propose a novel information theoretical framework for generalized multi-view learning. Guided by it, we build a multi-view coding method with a three-tier progressive architecture, namely Information theory-guided heuristic Progressive Multi-view Coding (IPMC). In the distribution-tier , IPMC aligns the distribution between views to reduce view-specific noise. In the set-tier , IPMC constructs self-adjusted contrasting pools, which are adaptively modified by a view filter. Lastly, in the instance-tier , we adopt a designed unified loss to learn representations and reduce the gradient interference. Theoretically and empirically, we demonstrate the superiority of IPMC over state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Unlike current state-of-the-art methods based on data augmentation and adversarial frameworks to solve the fair classification problem, this paper proposes the Information Bottleneck Disentanglement Based Sparse Representation algorithm to ensure the fairness of extracted features. The algorithm assumes that each sample is collaboratively generated by a categorical latent variable and a sensitive or category-irrelevant latent variable. To ensure that these two latent variables correctly capture their corresponding features, we impose Information Bottleneck Disentanglement constraints on them. Specifically, the Information Bottleneck is used to obtain more relevant information while compressing the amount of information. Simultaneously, a mutual information minimization constraint enhances their independence. Additionally, by employing the sparse structure, we can further disentangle the discriminative and sensitive information under supervision guidance. Therefore, the algorithm can improve the discrimination of the categorical latent variable as much as possible without extracting the sensitive data into it. Extensive experiments on five public and challenging benchmark datasets reveal the effectiveness of our proposed method, demonstrating that it can ensure fairness without sacrificing classification performance. • A new sparse representation method named IBD-SR is proposed for fair classification. • IBD-SR attempts to solve the fair classification problem based on information theory. • IBD-SR integrates sparse structures into information bottlenecks for the first time. • Experimental results validate the effectiveness and superiority of our IBD-SR method. [ABSTRACT FROM AUTHOR]

In the first part of this paper we develop some theorems in linear algebra applicable to information theory when all random variables involved are linear functions of the individual bits of a source of independent bits. We say that a collection of subspaces of a vector space are coordinated if the vector space has a basis such that each subspace is spanned by its intersection with the basis. We measure the failure of a collection of subspaces to be coordinated by an invariant that we call the discoordination of the family. We develop some foundational results regarding discoordination. In particular, these results give a number of new formulas involving three subspaces of a vector space. We then apply a number of our results, along with a method of Tian to obtain some new lower bounds in a special case of the basic coded caching problem. In terms of the usual notation for these problems, we show that for N = 3 documents and K = 3 caches, we have 6 M + 5 R ≥ 11 for a scheme that achieves the memory-rate pair (M , R) , assuming the scheme is linear. [ABSTRACT FROM AUTHOR]

The purpose of this paper is to demonstrate the use of Rough Set Theory to identify information that can be used in project management, namely the knowledge about the appropriate prioritization of activities and the use of resources directed towards the adopted project objectives. Rough Set Theory operates on large data sets and permits the reduction of irrelevant data and the induction of decision rules that discover recursion and relationships in the data. The discovered rules can prove valuable for decision-making project managers. All the rules found in course of the proposed rule-finding process have been considered interesting and are discussed in the paper. The paper also provides an example of the application of Rough Set Theory in the field mentioned above. [ABSTRACT FROM AUTHOR]

This paper presents an evaluation of the system-level integrated conceptual information of a major complex for a small-scale network containing two loops in accordance with the integrated information theory 3.0 framework. We focus on the following parameters characterizing the system model: (1) number of nodes in the loop, (2) frustration of the loop, and (3) temperature controlling the stochastic fluctuation of the state transition. Effects of these parameters on the integrated conceptual information and conditions for major complexes formed by a single loop, rather than the entire network, are investigated. Our first finding is that parity of the number of nodes forming a loop has a strong effect on the integrated conceptual information. For loops with an even number of nodes, the number of concepts tends to decrease, and the integrated conceptual information becomes smaller. Our second finding is that a major complex is more likely to be formed by a small number of nodes under small stochastic fluctuations. On the other hand, the entire network can easily become a major complex under larger stochastic fluctuations, and this tendency can be reinforced by frustration. It is also shown that, although counterintuitive, the integrated conceptual information can be maximized in the presence of stochastic fluctuations. These results suggest that even when several small subnetworks are connected by only a few connections, such as a bridge, the entire network may become a major complex by introducing some stochastic fluctuations and by frustrating loops with an even number of nodes. • Parity of the number of nodes in a loop strongly affects the integrated information. • Frustration in a loop can generate a major complex formed by the entire network. • Small stochastic fluctuations cause a major complex formed by a small number of nodes. • Moderate stochastic fluctuations can increase the integrated information. [ABSTRACT FROM AUTHOR]

Magnetic resonance-guided focused ultrasound (MRgFUS) lesioning of the ventralis intermedius nucleus (VIM) has shown promise in treating drug-refractory essential tremor (ET). It remains unknown whether focal VIM lesions by MRgFUS have broader restorative effects on information flow within the whole-brain network of ET patients. We applied an information-theoretical approach based on intrinsic ignition and the concept of transfer entropy (TE) to assess the spatiotemporal dynamics after VIM-MRgFUS. Eighteen ET patients (mean age 71.44 years) underwent repeated 3T resting-state functional magnetic resonance imaging combined with Clinical Rating Scale for Tremor (CRST) assessments one day before (T0) and one month (T1) and six months (T2) post-MRgFUS, respectively. We observed increased whole brain ignition-driven mean integration (IDMI) at T1 (p < 0.05), along with trend increases at T2. Further, constraining to motor network nodes, we identified significant increases in information-broadcasting (bilateral supplementary motor area (SMA) and left cerebellar lobule III) and information-receiving (right precentral gyrus) at T1. Remarkably, increased information-broadcasting in bilateral SMA was correlated with relative improvement of the CRST in the treated hand. In addition, causal TE-based effective connectivity (EC) at T1 showed an increase from right SMA to left cerebellar lobule crus II and from left cerebellar lobule III to right thalamus. In conclusion, results suggest a change in information transmission capacity in ET after MRgFUS and a shift towards a more integrated functional state with increased levels of global and directional information flow. • For the first time, functional consequences of a focal VIM-lesion, are described by advanced information theory-based measures. • Information broadcasting between hubs of the tremor network is improved post-MRgFUS, correlating with tremor improvement. • VIM-MRgFUS improves global information transfer, as evidenced by increased mean integration, suggesting restorative effects. [ABSTRACT FROM AUTHOR]

Although reinforcement learning (RL) can solve complex tasks after training, it's difficult to extend trained agents to target environments with perturbations. This lack of generalization ability has caused the dilemma of large-scale application of RL. Rich neural networks have brought huge improvements to deep learning, but these aren't adapted to RL and even bring negative effects. Therefore, RL algorithms have few choices of neural network, which greatly limits the representation and generalization ability of RL. To overcome these limitations, we propose a deep semi-dense compression network (DSCN) to improve RL generalization ability. First, we perform a structural extension on a general network model for RL. Then based on the information theory, we propose a semi-dense connection to enhance the neural information flow (NIF) of initial features. Finally, drawing on the ideas of course learning, we propose a channel compression approach to filter the redundant information of the deep network. In addition, we innovatively extend the experimental environment and evaluation metrics of the existing platform, which can fully evaluate the performance of DSCN. The experimental results show that our model achieves stable and significant generalization performance improvement. [ABSTRACT FROM AUTHOR]

Typical statistical methods of data analysis only handle determinate uncertainty, the type of uncertainty that can be modeled under the Bayesian or confidence theories of inference. An example of indeterminate uncertainty is uncertainty about whether the Bayesian theory or the frequentist theory is better suited to the problem at hand. Another example is uncertainty about how to modify a Bayesian model upon learning that its prior is inadequate. Both problems of indeterminate uncertainty have solutions under the proposed framework. The framework is based on an information-theoretic definition of an incoherence function to be minimized. It generalizes the principle of choosing an estimate that minimizes the reverse relative entropy between it and a previous posterior distribution such as a confidence distribution. The simplest form of the incoherence function, called the incoherence distribution, is a min-plus probability distribution, which is equivalent to a possibility distribution rather than a measure-theoretic probability distribution. A simple case of minimizing the incoherence leads to a generalization of minimizing relative entropy and thus of maximizing entropy. The framework of minimum incoherence is applied to problems of Bayesian-confidence uncertainty and to parallel problems of indeterminate uncertainty about model revision. [ABSTRACT FROM AUTHOR]

In this manuscript, we present a generalized deformed sum inspired by the nonadditive property of entropies such as those investigated by Tsallis and Shannon in the context of information theory. From this deformed sum, we define a generalization of the Vitali set and prove its nonmeasurability. Moreover, the standard sum is recovered as the deforming parameter tends to zero, and Vitali's theorem is retrieved. In particular, the present work is a generalization of the results derived in [2] for arbitrary functions satisfying general conditions, and not only nonextensive statistics. [ABSTRACT FROM AUTHOR]

• Imitation Learning with full state-action observations is not manageable in many applications. • Generalizing Imitation Learning to partial observations offers a natural setting for learning. • Facing the problem through Bayesian Inference yields a Behavioral Cloning strategy. • Recasting as an Information-Theoretic projection we can face multiple observation sequences. • Connection with Entropy Regularized Markov Decision Processes are investigated. In this work we formulate and treat an extension of the Imitation from Observations problem. Imitation from Observations is a generalisation of the well-known Imitation Learning problem where state-only demonstrations are considered. In our treatment we extend the scope of Imitation from Observations to feature-only demonstrations which could arguably be described as partial observations. Therewith we mean that the full state of the decision makers is unknown and imitation must take place on the basis of a limited set of features. We set out for methods that extract an executable policy directly from those features which, in the literature, would be referred to as Behavioural Cloning methods. Our treatment combines elements from probability and information theory and draws connections with entropy regularized Markov Decision Processes. [ABSTRACT FROM AUTHOR]

Spence's theory of signaling shows that signals can resolve adverse selections in job markets. Instead of signals, I consider what Spence calls indices and show that if the costs of knowing the indices are not nil, they could play the role of resolving adverse selection in sequential Bayesian games. I also show that by incorporating costs that are measured by the mutual information of indices, adverse selection can be resolved if the costs fall within a certain range. • Consider the market for lemons. • Indices are observable and unalterable attributes of the players. • Indices could play the role of resolving adverse selection. • The costs of indices are measured by their mutual information. • We could resolve adverse selection if the costs fell in a certain range. [ABSTRACT FROM AUTHOR]

• Presenting an adaptive scheme of estimating the potential of cascading failure as a result of transmission line outages through large power systems. • Developing a set of dominant operating variables to identify cascading failures caused by line outages. • Evaluating the line information consisting of dynamic securities and sensitivities in which the lines with the potential of cascading failures are identified online. • Developing mutual information theory based on the Least Square Error (LSE) method to estimate power system cascading failures without acting any line outages. In this study, the potential of cascading failure as a result of transmission line outages through large power systems is evaluated. Transmission lines are more dispose of the outage resulting from different fault occurrences and extend the uncontrolled failures compared to other power system devices. In this case, power system dynamic security must evaluate the possibility of cascading failures using a power system control center (PSCC) in order to limit the spread of outages caused by line failures. For this issue, using the PSCC online information consisting of power system operating variables, it is possible to present a variables-based scheme to estimate power system cascading failures concerning fault events. In order to cover the issue, based on developing a set of dominant operating variables (DOVs), this paper presents an online scheme of identifying cascading failures caused by line outages in the presence of line operational diversities and communication restrictions. In this scenario, during each time interval, by evaluating the proposed DOVs, the line information consisting of dynamic securities and sensitivities to cascading failure are investigated which the lines with the potential of cascading failures are identified online. Proper DOVs are determined for this issue by studying a variety of mathematical formulations according to theory of mutual information (MI) and measuring the entropy among the variables. In a real-time environment, by using the proposed DOVs and substituting them through online boundary equations based on the Least Square Error (LSE) method, the potentials of power system cascading failures without acting any line outage are provided. The effectiveness of the suggested technique is evaluated using a typical IEEE-39 bus and IEEE-118 bus, and proper results with high accuracy are achieved by evaluating the system potential concerning large cascading failures. [ABSTRACT FROM AUTHOR]

This paper revisits classical work of Rauch et al. (1965) and develops a novel statistical method for maximum likelihood (ML) recursive state estimation in general state–space models. The new method is based on statistical estimation theory for incomplete information, which has been well developed primarily for ML parameter estimation (Dempster et al., 1977). Distributional identities for the posterior score function and information matrix of state are established. Using these identities, a fast convergent EM-gradient algorithm is proposed, extending the Lange (1995) algorithm for ML recursive state estimation. It revisits and provides an improvement to the EM-algorithm of Ramadan and Bitmead (2022). An explicit form of the information matrix is developed to provide empirical estimates of the standard errors. Sequential Monte Carlo method is used for the valuation of the score function, information and posterior covariance matrices. Some numerical examples are discussed to exemplify the main results. [ABSTRACT FROM AUTHOR]

Ships are unlikely to share their navigation plans at sea, and officers on watch (OOWs) iteratively make decisions to avoid collisions based on limited information and predicted intention. Therefore, ship-ship collision avoidance can be considered a dynamic imperfect information game. Existing conflict risk assessment methods have focused on the uncertainty of ship motion and did not consider game behavior and the bounded rationality of OOWs. This paper proposes a novel dynamic risk assessment method based on the quantal response equilibrium, a modified game-theory model with imperfect information. Expected utility function is employed to quantify the behavior (yielding or not yielding) benefits, incorporating objective risk, energy loss and predicted intention. The OOWs' behavior probability is predicted by considering perception and intention uncertainties. We develop three stages to describe the OOWs' decisions in different risk zones from conflict formation to the end of collision avoidance. Dynamic conflict risk is estimated by calculating the joint probability that OOWs do not yield when encountering ships in the three stages based on the game theory. Simulations and a case study demonstrate that the method realistically reflects conflict risk and reveals ship characteristics influencing yielding. The results provide information for maritime traffic management and OOW training. • Dynamic conflict risk assessment method considering game behaviors of OOWs is proposed. • The bound rationality of OOWs due to perception and intention uncertainties is investigated. • Game behaviors are modeled based on imperfect information game theory and expected utility theory. • A conflict probability prediction method is proposed based on three-stage dynamic game. • The method obtained more practical results verified by simulation and real case. [ABSTRACT FROM AUTHOR]

Granular structures are mathematical representations of knowledge used in granular computing. As a new type of granular structure, fuzzy β -covering has attracted widespread attention in recent years. One of the most critical related studies is measuring the uncertainty of fuzzy β -coverings. It is the foundation for studying fuzzy β -covering applications, such as classification and clustering. In this paper, we investigate the uncertainty measures of fuzzy β -coverings from the viewpoints of algebra and information theory, and propose the corresponding methods of fuzzy β -covering reduction for feature selection. Firstly, a generalized fuzzy β -neighborhood and the associated fuzzy β -covering rough sets are presented. Secondly, three types of accuracy measures of fuzzy β -coverings are constructed via fuzzy β -covering rough sets. On this basis, the self-information of fuzzy β -coverings is defined by fusing accuracy and roughness measures. Based on the proposed uncertainty measures, two heuristic methods of fuzzy β -covering reduction are put forward and the corresponding algorithms are designed for monotonic feature selection. Finally, the performance of the proposed methods is compared with several mainstream feature selection methods and experimental results demonstrate the rationality and superiority of our proposed methods. • The concept of generalized fuzzy β -neighborhood is introduced. • Multi-fuzzy β -covering fusion-based accuracy and self-information are proposed. • Fuzzy dependency function and fuzzy self-information are defined. • Two monotonic fuzzy β -covering reduction methods are proposed. [ABSTRACT FROM AUTHOR]

• This paper proposed a novel diagnostic model called Non-Contact Sensing Data Fusion Driven Cross-Modal Zero-Sample Diagnostic Network. • This model leverages the principles of information theory, specifically focusing on information fusion theory. • This approach enhances zero-sample diagnostic performance through efficient information utilization and cross-modal data fusion. • This strategy effectively reduces misclassification risks and augments the robustness of the proposed zero-sample diagnostic network against interference from load variations. Gearboxes, fundamental components in the domains of manufacturing, transportation, and aerospace apparatus, are highly susceptible to impairments. The emerging technique of non-contact sensing measurement holds promise as a valuable tool for real-time monitoring and diagnosis of gearboxes, especially in challenging, dynamic environments. However, historical non-contact diagnostic techniques often face difficulties due to load-variation-induced interference, which leads to inconsistent diagnostic results. Moreover, acquiring sufficient non-contact sensing data for specific rare or severe gearbox faults can be exceedingly challenging and, at times, unattainable. This data scarcity limits the application of existing supervised and semi-supervised non-contact diagnostic methods. To address these limitations, this paper develops a novel diagnostic model, called Non-Contact Sensing Data Fusion Driven Cross-Modal Zero-sample Diagnostic Network. Specifically, we first develop a novel central fusion module to facilitate feature-level fusion at both global and local scales utilizing infrared thermal and acoustic data, thereby establishing a rich and comprehensive fault representation. Following that, we construct an end-to-end cross-modal zero-sample diagnostic architecture to mine and fuse complementary fault information. This approach enhances zero-sample diagnostic performance through efficient information utilization and cross-modal feature fusion. Afterward, we adopt a cross-modal self-enhancing fusion strategy during the training phase, which improves the information fusion performance of intra-class features from various modalities. This strategy effectively reduces misclassification risks and augments the robustness of the proposed zero-sample diagnostic network against interference from load variations. Comprehensive experimental results confirm that our proposed NCFZD sets the new state-of-the-art in multiple non-contact, zero-sample diagnostic scenarios, reaching HM scores of 86.06 %, 81.53 %, and 62.56 % respectively. By incorporating the information fusion theory, this model advances gearbox diagnostics in non-contact sensing and contributes to the broader field of information fusion theory by demonstrating its practical application in real-world problem-solving scenarios. [ABSTRACT FROM AUTHOR]

Information granulation theory has been widely used in short-term time-series forecasting research and holds significant weight. However, the error accumulation due to the lack of granular accuracy, along with information redundancy or deficiency in predictions, significantly affects short-term prediction accuracy. To compensate for these shortcomings, this paper proposes a double-level optimal fuzzy association rules prediction model for short-term time-series forecasting, which can strengthen the performance of information granulation in prediction. Firstly, this paper proposes a concept of breakpoints, which can accurately segment complex linear trends in time series and thus obtain a granular time series with highly accurate linear fuzzy information granules (LFIGs). Secondly, a improved distance is proposed to more accurately reflect the similarity between LFIGs by addressing counter-intuitive problems in the original distance. Theoretical analysis shows that the improved distance can effectively reduce errors in granular calculation. Then, a granule-suited fuzzy c-means algorithm is proposed for clustering LFIGs. Finally, this paper proposes a double-level optimal fuzzy association rules prediction model, which establishes the optimal rules for each cluster and selects the optimal two rules for prediction by the contribution of the clusters. The experimental results show that the prediction method effectively avoids the problems of information redundancy and information deficiency, and increases forecast accuracy. The model's exceptional performance is demonstrated through comparative analysis with existing models in experimental investigations. • A new concept of breakpoints is proposed to partition time series linear trends. • A DTW-iL1 distance is proposed to measure the similarity between granules. • DTW-iL1 Fuzzy C-Means(DFCM) is proposed for clustering granules. • Propose double-level optimal fuzzy association rules prediction model based on DFCM. [ABSTRACT FROM AUTHOR]

This paper revisits the empirical relationship between interest rates and money demand from a novel perspective, i.e., information theory. Particularly, we utilize the model-free transfer entropy to quantify the flow of information from interest rates to monetary aggregates and present three findings. First, we document a hump-shaped informational link between interest rate and M1 monetary aggregate, with a rounded high point in the late 1980s and early 1990s. Second, we identify three structural shifts in the information transmission from interest rate to M1. The first two breakpoints occurred in the early 1980s and mid-1990s, likely as a response to the removal of Regulation Q and the introduction of sweep technology, respectively. The third shift took place during the relatively less-explored period of the early 2000s. Finally, we unravel a previously unreported pivotal distinction between the first two changepoints despite the apparent similarity in inducing money demand instability: the 1980s financial deregulations facilitate the transmission of information, whereas the 1990s financial reforms acted as an impediment to the information flow. Our results are robust to alternative entropy measures. [ABSTRACT FROM AUTHOR]

We reformulate models in epidemiology and population dynamics in terms of probability distributions. This allows us to construct the Fisher information, which we interpret as the metric of a one-dimensional differentiable manifold. For systems that can be effectively described by a single degree of freedom, we show that their time evolution is fully captured by this metric. In this way, we discover universal features across seemingly very different models. This further motivates a reorganisation of the dynamics around zeroes of the Fisher metric, corresponding to extrema of the probability distribution. Concretely, we propose a simple form of the metric for which we can analytically solve the dynamics of the system that well approximates the time evolution of various established models in epidemiology and population dynamics, thus providing a unifying framework. [ABSTRACT FROM AUTHOR]

Among the Graph Neural Network (GNN) models that address the task of node classification, Graph Echo State Networks (GESN) have proved particularly effective in addressing the challenge of heterophily, i.e. the presence of a significant fraction of inter-class edges in the learning task graph. The effectiveness of GESN is paired with its efficiency, owing to the reservoir computing paradigm. While previous literature has analyzed the design of reservoirs for sequence ESN and GESN for graph-level tasks, the problem of providing effective designs of reservoirs for node-level GESN is so far largely unexplored. In this paper we analyze the impact of different reservoir designs on node classification accuracy and on the quality of node embeddings computed by GESN, focusing both on dense and sparse reservoir layouts. As measures of embedding richness, we adopt both graph topology-dependent metrics previously employed in the analysis of embedding smoothing, and topology-independent metrics from the areas of information theory and numerical analysis. In particular, we propose the application of entropy measures for quantifying information in node embeddings. [Display omitted] • First systematic analysis of reservoir computing designs for node classification. • Proposing a set of topology-aware and -agnostic richness measures for node embeddings. • Random ensembles and dense orthogonal reservoirs perform generally better. • Extremely sparse reservoirs preserve accuracy and node embedding quality. [ABSTRACT FROM AUTHOR]

The gauge formulation of Einstein gravity in AdS 3 background leads to a boundary theory that breaks modular symmetry and loses the covariant form. We examine the Weyl anomaly for the cylinder and torus manifolds. The divergent term is the same as the Liouville theory when transforming from the cylinder to the sphere. The general Weyl transformation on the torus also reproduces the Liouville theory. The Weyl transformation introduces an additional boundary term for reproducing the Liouville theory, which allows the use of CFT techniques to analyze the theory. The torus partition function in this boundary theory is one-loop exact, and an analytical solution to disjoint two-interval Rényi-2 mutual information can be obtained. We also discuss a first-order phase transition for the separation length of two intervals, which occurs at the classical level but is smoothed out by non-perturbative effects captured by averaging over a modular group in the boundary theory. [ABSTRACT FROM AUTHOR]

Feature selection is a popular preprocessing step to reduce the dimensionality of the data while preserving the important information. In this paper, we propose an efficient and green feature selection method based on information theory, with the novelty of using the logarithmic division and resorting to fixed-point precision. Moreover, we extend these advancements by adapting the Mutual Information calculation for federated scenarios. The results of experiments conducted on several datasets indicate the potential of our proposal, as it does not incur significant information loss compared to the double-precision method, both in the features selected and in the subsequent classification step. Our method has shown significant potential when applied in federated scenarios, where experimentation demonstrated a lossless feature selection process and maintains classification results compared with centralised versions. These findings open up possibilities towards a new family of green feature selection methods, which would help to minimise energy consumption, lower carbon emissions and increase adaptability to Internet of Things environments. • Implementation of green feature selection methods based on information theory. • Study of logarithmic division to reduce energy and memory consumption. • Federated Mutual Information calculation enables IoT environments data privacy. • Feature selection with logarithmic division obtains similar results as double-precision. • Lossless and robust methods have been developed for federated scenarios. [ABSTRACT FROM AUTHOR]

Computed tomography (CT) based lung screening followed by computer-vision-based automated segmentation has great potential in early diagnosis of COVID-19, but low sample size and high patient-to-patient variability of infection characteristics are two of the major challenges. We propose a novel Mutual Information Balanced Net (MIB-Net) with three distinctive features. First, it captures mutual information across feature maps, CT scans, and predicted segmentation labels to optimize the contribution of different feature maps for better segmentation accuracy. Second, it incorporates a three-dimensional structural prior of CT scans based on predicted segmentation of neighboring CT slices to improve accuracy. Third, a novel Weight Optimization Filter (WOF) is used to adjust the proportion of 3D structural prior to be incorporated, which further improves the segmentation accuracy. Experimental results show that the proposed approach outperforms state-of-the-art biomedical image segmentation models by 7.87%, 3.85%, and 5.01% in Dice Score, respectively for three popular COVID-19 CT datasets. The better performance of the proposed approach suggests that it may serve as an attractive alternative to Lung CT-based screening for COVID-19. • Incorporation of Mutual Information to control the parallel information flow. • Novel MI Balancer architecture for COVID-19 segmentation. • Incorporation of the 3D structural prior to learn inter-slice spatial dependence. • Novel WOF to balance between the current slice and neighboring prediction. [ABSTRACT FROM AUTHOR]

Digital Engineering (DE) must be able to maximize efficiency and update dynamically in order to keep up with the wave of global interconnection. The evolution of the Digital Twin (DT) plays a significant role in DE as it enables autonomous optimization and an iterative cycle. However, the intricacy of real-world settings renders conventional evolutionary investigations of a solitary DT inadequate in tackling intricate systems at a comprehensive lifecycle magnitude. To address this, this paper presents the Co-evolutionary DTs (CoEDT), which has several key characteristics. (1) CoEDT study the evolutionary behavior of interconnected multi-DT. (2) CoEDT use a co-evolutionary distributed system architecture and is a DT technology that integrate MBSE. (3) CoEDT offers detailed dynamic models for the complex interactions and co-evolution among multi-DT throughout the lifecycle. It also supports real time measurement of multi-DT behavior to detect system anomalies. The significance of CoEDT lies in providing a more comprehensive insight for future product development by constructing a parallel world that simulates the lifecycle. In the CoEDT, we tackle the behavioral identification and structure of these evolving multi-DT throughout product lifecycle through the following approaches. Initially, drawing inspiration from biological cytology, we have conceived a concept of Collective DTs (CollDTs) to observing a collective behavioral of multi-DT. Subsequently, we further developed CoEDT to depict co-evolutionary behavior patterns and established a co-evolution architecture for all DT throughout the lifecycle by fusing Model-Based Systems Engineering (MBSE). Then, dynamic expressions and algorithm that can measure the co-evolution of every DT are inferred using information theory. Finally, the viability of the proposed CoEDT framework is demonstrated through the development of a solid rocket engine, which promotes the application of DT in the DE. [ABSTRACT FROM AUTHOR]

Machine Learning (ML) and Algorithmic Information Theory (AIT) look at Complexity from different points of view. We explore the interface between AIT and Kernel Methods (that are prevalent in ML) by adopting an AIT perspective on the problem of learning kernels from data, in kernel ridge regression, through the method of Sparse Kernel Flows. In particular, by looking at the differences and commonalities between Minimal Description Length (MDL) and Regularization in Machine Learning (RML), we prove that the method of Sparse Kernel Flows is the natural approach to adopt to learn kernels from data. This approach aligns naturally with the MDL principle, offering a more robust theoretical basis than the existing reliance on cross-validation. The study reveals that deriving Sparse Kernel Flows does not require a statistical approach; instead, one can directly engage with code-lengths and complexities, concepts central to AIT. Thereby, this approach opens the door to reformulating algorithms in machine learning using tools from AIT, with the aim of providing them a more solid theoretical foundation. [ABSTRACT FROM AUTHOR]

Probabilistic control design is founded on the principle that a rational agent attempts to match modelled with an arbitrary desired closed-loop system trajectory density. The framework was originally proposed as a tractable alternative to traditional optimal control design, parametrizing desired behaviour through fictitious transition and policy densities and using the information projection as a proximity measure. In this work we introduce an alternative parametrization of desired closed-loop behaviour and explore alternative proximity measures between densities. It is then illustrated how the associated probabilistic control problems solve into uncertain or probabilistic policies. Our main result is to show that the probabilistic control objectives majorize conventional, stochastic and risk sensitive, optimal control objectives. This observation allows us to identify two probabilistic fixed point iterations that converge to the deterministic optimal control policies establishing an explicit connection between either formulations. Further we demonstrate that the risk sensitive optimal control formulation is also technically equivalent to a Maximum Likelihood estimation problem on a probabilistic graph model where the notion of costs is directly encoded into the model. The associated treatment of the estimation problem is then shown to coincide with the moment projected probabilistic control formulation. That way optimal decision making can be reformulated as an iterative inference problem. Based on these insights we discuss directions for algorithmic development. [ABSTRACT FROM AUTHOR]

Regressions, or backward saccades, are common during reading, accounting for between 5% and 20% of all saccades. And yet, relatively little is known about what causes them. We provide an information-theoretic operationalization for two previous qualitative hypotheses about regressions, which we dub reactivation and reanalysis. We argue that these hypotheses make different predictions about the pointwise mutual information or pmi between a regression's source and target. Intuitively, the pmi between two words measures how much more (or less) likely one word is to be present given the other. On one hand, the reactivation hypothesis predicts that regressions occur between words that are associated, implying high positive values of pmi. On the other hand, the reanalysis hypothesis predicts that regressions should occur between words that are not associated with each other, implying negative, low values of pmi. As a second theoretical contribution, we expand on previous theories by considering not only pmi but also expected values of pmi , E [ pmi ], where the expectation is taken over all possible realizations of the regression's target. The rationale for this is that language processing involves making inferences under uncertainty, and readers may be uncertain about what they have read, especially if a previous word was skipped. To test both theories, we use contemporary language models to estimate pmi -based statistics over word pairs in three corpora of eye tracking data in English, as well as in six languages across three language families (Indo-European, Uralic, and Turkic). Our results are consistent across languages and models tested: Positive values of pmi and E [ pmi ] consistently help to predict the patterns of regressions during reading, whereas negative values of pmi and E [ pmi ] do not. Our information-theoretic interpretation increases the predictive scope of both theories and our studies present the first systematic crosslinguistic analysis of regressions in the literature. Our results support the reactivation hypothesis and, more broadly, they expand the number of language processing behaviors that can be linked to information-theoretic principles. [ABSTRACT FROM AUTHOR]

• We consider supply chain visibility a crucial resource for digital technologies. • We found that digital technologies significantly enhance the resilience of healthcare supply chains in highly dynamic environments. • Our findings contribute to the organisational information processing theory. • Our findings offer numerous directions to healthcare organisations and policymakers to exploit digital technologies in a highly turbulent environment. The healthcare sector is a complex and critical industry that requires a resilient supply chain to ensure patients receive the care they need. However, the pandemic and geopolitical crises have highlighted the need for a more comprehensive understanding of healthcare supply chain resilience. To address this, healthcare providers and policymakers increasingly recognise the importance of digital technology in enhancing their supply chain resilience. Despite this recognition, there is a lack of empirical studies that have explored the relationship between supply chain visibility, digital capabilities, and environmental dynamism in the healthcare sector. Therefore, our study aimed to fill this gap by utilising organisational information processing theory (OIPT) to explain the relationship between these factors. Our analysis of 137 survey responses revealed that supply chain visibility is a foundational resource for digital capabilities. Moreover, digital capabilities are vital in enhancing healthcare supply chain resilience in highly turbulent environments. Our research expands the concept of the OIPT to offer a more comprehensive explanation of how supply chain visibility is a crucial factor for the effective functioning of digital technologies, particularly in highly uncertain environments. Our findings also provide useful insights for healthcare managers and policymakers who wish to enhance their supply chain's resilience and leverage digital technologies to achieve this objective. [ABSTRACT FROM AUTHOR]

We introduce an innovative and mathematically rigorous definition for computing common information from multi-view data, drawing inspiration from Gács-Körner common information in information theory. Leveraging this definition, we develop a novel supervised multi-view learning framework to capture both common and unique information. By explicitly minimizing a total correlation term, the extracted common information and the unique information from each view are forced to be independent of each other, which, in turn, theoretically guarantees the effectiveness of our framework. To estimate information-theoretic quantities, our framework employs matrix-based Rényi's α -order entropy functional, which forgoes the need for variational approximation and distributional estimation in high-dimensional space. Theoretical proof is provided that our framework can faithfully discover both common and unique information from multi-view data. Experiments on synthetic and seven benchmark real-world datasets demonstrate the superior performance of our proposed framework over state-of-the-art approaches. • We formulate a mathematically definition of common information for multiview data. • We propose a multiview learning method that discerns common and unique information. • Our method is scalable to data more than two view, supported by theoretical proof. • Experimental results substantiate the efficacy of the proposed framework. [ABSTRACT FROM AUTHOR]

Guiding tourists in displaying civilized behavior is pivotal in destination management. This research tries to understand what forms of interpretation can effectively promote tourists' desirable behaviors. Grounded in social information processing theory, it introduces the tool of anthropomorphic interpretation cues and investigates its' efficacy in shaping tourists' civilized behavioral intentions through meaningful experience. Hypotheses are developed by integrating the social information processing theory, the three-factor theory of anthropomorphism and the narrative transportation theory, in three experiments. First, we confirm how anthropomorphic interpretation cues substantially enhance civilized behavioral intentions. Second, we show that the perception of meaningful experience mediates the effect of anthropomorphic interpretation cues in interpretive communication. Third, we show that the presence (vs. absence) of narrative within anthropomorphic interpretation cues amplifies the intervention effect. These findings offer valuable insights for destination managers, suggesting the strategic implementation of anthropomorphic cues and narrative approaches in interpretation to encourage tourists' civilized behavior. [Display omitted] • This study focuses on the impact of anthropomorphic interpretation cues on tourists civilized behavioral intentions. • Anthropomorphic interpretation cues can enhance tourists civilized behavioral intentions. • Meaningful experience mediates the effect of anthropomorphic interpretation cues on tourists civilized behavioral intentions. • Narrative strengthens the anthropomorphic interpretation cues on tourists civilized behavioral intentions. [ABSTRACT FROM AUTHOR]

Data-driven synthesis of human motion during conversational speech is an active research area with applications that include character animation, computer gaming and conversational agents. Natural looking motion is key to both perceived realism and understanding of any synthesised animation. Multi-modal speech and body-motion data is scarce and limited, so it is common to augment real motion data by mirroring the body pose to double the number of training samples. This augmentation is based on the assumption that a person's gesturing is not affected by handedness and that the reflected pose is plausible. In this study, we explore the validity of this assumption by evaluating the reflective symmetry of a speaker's arms during conversational exchanges. We analyse the left and right arm motion of 36 subjects during dyadic conversation and present the per-frame symmetry of the arm gestures. To identify temporal offsets caused by the presence of a leading hand, we compute the time lag between movements of the left and right arms. We perform a nearest neighbour search to test the validity of any mirrored pose. We also consider information theory to examine the information gain from mirroring the data. We implement a speech-to-gesture generative model to determine the efficacy of lateral mirroring techniques for data augmentation. Our findings suggest that both positional symmetry and left–right motion offsets vary from speaker to speaker. We conclude that data augmentation by mirroring is valid in certain cases when considering the mirrored pose as a new virtual identity, but that it should be carefully considered as a generic approach if the gesturing style and handedness of the original speaker is to be maintained. • Review the motion symmetry of multiple speakers during dyadic conversation. • Analyse positional, temporal and informational symmetry of arm motion. • Discuss the efficacy of lateral mirroring of the human body for data augmentation. • Conclude lateral mirroring is not suited as a generic approach. • Suggest lateral mirroring as a new identity is a suitable data augmentation method. • Propose our statistical analysis for evaluating speech-driven conversational agents. [ABSTRACT FROM AUTHOR]

The non-uniform distribution of assembly features such as form error and stress has a significant influence on the accuracy and stability of precision instruments. Therefore, to improve assembly precision, non-uniformly distributed assembly features must be evaluated in terms of value non-uniformity as well as location non-uniformity. However, the current effective non-uniformity evaluation methods for assembly features only consider the value non-uniformity without considering the influence of the data location. This study proposes a multidimensional entropy evaluation method (MDEEM) and establishes multidimensional entropy evaluation indexes (MDEEIs) to evaluate the non-uniformity of assembly features in a plane region considering both their value and location. First, the mathematical model and MDEEIs for non-uniform data in a one-dimensional (1D) geometric space were developed and verified, wherein the concept of control distance was introduced and the information entropy theory was adopted. Then, the MDEEIs for non-uniform data in a two-dimensional (2D) geometric space were established and verified based on those in the 1D space. Subsequently, the MDEEM was proposed to demonstrate the non-uniformity evaluation process for assembly features in a 2D plane region. Finally, the proposed MDEEM was applied to reveal the distribution characteristics of two types of assembly features that mainly affect the assembly quality, namely form error and stress. This study provides an evaluation method for the non-uniform distribution of various assembly factors in a plane region and aids in the optimization of precision assembly processes. [Display omitted] • Mathematical models and evaluation indexes for 1D and 2D data are developed. • An evaluation method for non-uniform assembly features is established. • The proposed evaluation method and evaluation indexes are verified. • Distributions of machine-induced form error and assembly-induced stress are revealed. [ABSTRACT FROM AUTHOR]

Motivated by algorithmic information theory, the problem of program discovery can help find candidates of underlying generative mechanisms of natural and artificial phenomena. The uncomputability of such inverse problem, however, significantly restricts a wider application of exhaustive methods. Here we present a proof of concept of an approach based on IMP, a high-level imperative programming language. Its main advantage is that conceptually complex computational routines are more succinctly expressed, unlike lower-level models such as Turing machines or cellular automata. We investigate if a more expressive higher-level programming language can be more efficient at generating approximations to algorithmic complexity of recursive functions. • We explore an inverse problem for model discovery based on principles of algorithmic information. • The method can help find candidates of underlying generative mechanisms and first principles. • A proof of concept based on IMP, a high-level imperative programming language, is provided. • IMP expressions are succinct, unlike lower-level models such as formal automata. • We investigate if a more expressive higher-level programming language can be more efficient. [ABSTRACT FROM AUTHOR]

Algorithmic bias indicates the discrimination caused by algorithms, which occurs with protected features such as gender and race. Many researchers have tried to define the fairness and devise methods to mitigate bias, but it is still premature to reach the unanimous definition and evaluation metrics of fairness according to society, times and cultures. In this paper, we introduce three evaluation metrics such as parity difference, equalized opportunity and equalized odds that can deal with various definitions of algorithmic bias, and concretely divide the three general approaches further into seven methods with some challenges, resulting in relabeling, generation, fair representation (for pre-processing), constraint optimization, regularization (for in-processing), calibration and thresholding (for post-processing). Among them, the pre-processing method is widely used due to its versatility, but it has limitation to deal with the information on data and features related with bias appropriately. In order to preserve the characteristics of the original data while excluding the information about the features causing bias, we propose a preprocessing approach based on information theory that avoids collision in the dual optimization, where the latent space is divided into two subspaces. Experiments are conducted with the well-known benchmark datasets of Census and COMPAS, and two real-world tasks: facial emotion recognition and text sentiment analysis. The information theoretic approach is promising to achieve fair machine learning by reducing the bias caused by several features such as age, race and gender. [ABSTRACT FROM AUTHOR]