Search

In this paper, a novel approach to the self-organization of hierarchical prototype-based classifiers from data is proposed. The approach recursively partitions the data at multiple levels of granularity into shape-free clusters of different sizes, resembling Voronoi tessellation, and naturally aggregates the resulting cluster medoids into a multi-layered prototype-based structure according to their descriptive abilities. Different from conventional classification models, it is nonparametric and entirely data-driven, and the learned model can offer a high-level of transparency and interpretability thanks to the underlying prototype-based nature. The system identification process underpinning the approach is driven by the aim of separating data samples of different classes into nonoverlapping multi-granular clusters. Its associated decision-making process follows the “nearest prototype” principle and hence, the rationales of the subsequent decisions made can be explicitly explained. Experimental studies based on popular benchmark classification problems, as well as on a practical application to remote sensing image classification, demonstrate the efficacy of the proposed approach.

The file attached to this record is the author's final peer reviewed version. The Publisher's final version can be found by following the DOI link. In allusion to heterogeneous multi-criteria group decision making (MCGDM) problems with incomplete weights and q-rung orthopair fuzzy (q-ROF) truth degrees, where many kinds of criteria values, i.e., crisp values, intervals, trapezoidal fuzzy values, hesitant fuzzy values and q-ROF values (q-ROFVs), and multiple types of interactions exist, i.e., positive synergetic interactions, negative synergetic interactions and independence, a novel multi-objective q-ROF programming approach is proposed. In particular, in order to globally capture the interactions among criteria, Choquet-based relative closeness degrees are developed based on the technique for order performance by similarity to ideal solution (TOPSIS) and the Choquet integral. Then, the q-ROF Choquet-based group consistency index (q-ROFCGCI) and the q-ROF Choquet-based group inconsistency index (q-ROFCGII) are defined. Next, to derive optimal 2-additive fuzzy measures on the criteria set and optimal experts' weights, a new multi-objective q-ROF mathematical programming model is established by minimizing the q-ROFCGII and maximizing the q-ROFCGCI. Subsequently, an algorithm based on the adaptive non-dominated sorting genetic algorithm III (A-NSGA-III) is designed to solve the established model. Afterwards, the Choquet-based overall relative closeness degrees of the alternatives is used to obtain their preferred ordering. Finally, the effectiveness and advantage of the proposed approach is verified using four real cases concerning the evaluation of social commerce.

Iron tailings are an excellent secondary resource. Reasonable use, it can not only alleviate the problem of resource shortage, but also solve some environmental pollution problems. Therefore, it has a good development prospect. The resource utilization of iron tailings has always been a hot spot of social concern. The article reviews the results achieved by the predecessors. In addition, outlines the method of activation of iron tailings and the influence of its strength and durability as a concrete composite admixture. At last, the prospects of the research on the utilization of iron tailings are put forward.

In order to realize the energy efficient on the traffic and illuminating lights in city, the Intelligent Traffic and Illuminating Control System is proposed in this paper. The working criteria of real-time monitoring and intelligently control are designed. The key technologies including Internet of Things with power line communication by LonWorks technology and the piezoelectric sensor on site detecting are utilized in the research. The traffic density is detected by piezoelectric sensor and is "told" to LonWorks network to adjust the traffic red lamp lighting time and the illuminating time. The working flow of the Intelligent Traffic and Illuminating Control System has been verified to ensure the sufficient capability by Cp 1.72 and Cpk 1.60 within Statistical Process Control. The Energy efficient is achieved by 35.71% electrical energy reduction and 66.67% fuel consumption reduction on the energy saving and the carbon oxide emission reduction, which demonstrates that the Intelligent Control System achieves effective and valuable real-time monitoring and intelligently control with superiority of the energy efficient compared to the previous reports.

Based on the density-functional theory, the structural transitions manipulated by the biaxial strains are investigated in BaTiO3 film using the first-principles calculations. It is found that when the compressive strain is increased continuously in the single domain BaTiO3 films, a phase transition occurs from r phase to c phase, on the contrary, when the tensile strain is increased, there will be a successive phase transition from t phase to r phase, then to aa phase. All structural phase transitions are second-order, and r phase always integrates into other phases after phase transitions. Furthermore, it is found that the spontaneous polarization vector P → can rotates continuously in the 1 ¯ 10 plane, and fails to mutate during phase transition. At last, we also explain this strain induced polarization rotation in detail according to the orbital hybridization between Ti-O ions.

In this paper, we introduce a new form of describing fuzzy sets (FSs) and a new form of fuzzy rule-based (FRB) systems, namely, empirical fuzzy sets (εFSs) and empirical fuzzy rule-based (εFRB) systems. Traditionally, the membership functions (MFs), which are the key mathematical representation of FSs, are designed subjectively or extracted from the data by clustering projections or defined subjectively. εFSs, on the contrary, are described by the empirically derived membership functions (εMFs). The new proposal made in this paper is based on the recently introduced Empirical Data Analytics (EDA) computational framework and is closely linked with the density of the data. This allows to keep and improve the link between the objective data and the subjective labels, linguistic terms and classes definition. Furthermore, εFSs can deal with heterogeneous data combining categorical with continuous and/or discrete data in a natural way. εFRB systems can be extracted from data including data streams and can have dynamically evolving structure. However, they can also be used as a tool to represent expert knowledge. The main difference with the traditional FSs and FRB systems is that the expert does not need to define the MF per variable; instead, possibly multimodal, densities will be extracted automatically from the data and used as εMFs in a vector form for all numerical variables. This is done in a seamless way whereby the human involvement is only required to label the classes and linguistic terms. Moreover, even this intervention is optional. Thus, the proposed new approach to define and design the FSs and FRB systems puts the human “in the driving seat”. Instead of asking experts to define features and MFs correspondingly, to parameterize them, to define algorithm parameters, to choose types of MFs or to label each individual item, it only requires (optionally) to select prototypes from data and (again, optionally) to label them. Numerical examples as well as a naïve empirical fuzzy (εF) classifier are presented with an illustrative purpose. Due to the very fundamental nature of the proposal it can have a very wide area of applications resulting in a series of new algorithms such as εF classifiers, εF predictors, εF controllers, etc. This is left for the future research.

Machine learning, as a subarea of artificial intelligence, is widely believed to reshape the human world in the coming decades. This thesis is focused on both the unsupervised and supervised self-organising transparent machine learning techniques. One particularly interesting aspect is the transparent self-organising deep learning systems. Traditional data analysis approaches and most of the machine learning algorithms are built upon the basis of probability theory and statistics. The solid mathematical foundation of the probability theory and statistics guarantees the good properties of these learning algorithms when the amount of data tends to infinity and all the data comes from the same distribution. However, the prior assumptions of the random nature and same distribution imposed on the data generation model are often too strong and impractical in real applications. Moreover, traditional machine learning algorithms also require a number of free parameters to be predefined. However, without any prior knowledge of the problem, which is often the case in real situations, the performance of the algorithms can be largely influenced by the improper choice. Deep learning-based approaches are currently the state-of-the-art techniques in the fields of machine learning and computer vision. However, they are also suffering from a number of deficiencies including the computational burden of training using huge amount of data, lack of transparency and interpretation, ad hoc decisions about the internal structure, no proven convergence for the adaptive versions that rely on reinforcement learning, limited parallelisation and offline training, etc. These shortcomings largely all hinder the wider applications of the deep learning in real situations. The novel approaches presented in this thesis are developed within the Empirical Data Analytics framework, which is an alternative, but more advanced computational methodology to the traditional approaches based on the ensemble properties and mutual distribution of the empirical discrete observations. The novel self-organising transparent machine learning algorithms presented in this work for clustering, regression, classification and anomaly detection are autonomous, self-organising, data-driven and free from user- and problem- specific parameters. They do not impose any data generation models on the data a priori, but are driven by the empirically observed data and are able to produce the objective results without prior knowledge of the problems. In addition, they are highly efficient and suitable for large-scale static/streaming data processing. The newly proposed self-organising transparent deep learning systems are able to achieve human-level performance comparable to or even better than the deep convolutional neural networks on image classification problems with the merits of being fully transparent, self-evolving, highly efficient, parallelisable and human-interpretable. More importantly, the proposed deep learning systems have the ability of starting classification from the very first image of each class in the same way as humans do. Numerical examples based on numerous challenging benchmark problems and comparisons conducted with the state-of-the-art approaches presented in this thesis demonstrated the validity and effectiveness of the proposed new machine learning algorithms and deep learning systems and show their potential for real applications.

In this paper, we propose an approach to data analysis, which is based entirely on the empirical observations of discrete data samples and the relative proximity of these points in the data space. At the core of the proposed new approach is the typicality—an empirically derived quantity that resembles probability. This nonparametric measure is a normalized form of the square centrality (centrality is a measure of closeness used in graph theory). It is also closely linked to the cumulative proximity and eccentricity (a measure of the tail of the distributions that is very useful for anomaly detection and analysis of extreme values). In this paper, we introduce and study two types of typicality, namely its local and global versions. The local typicality resembles the well-known probability density function (pdf), probability mass function, and fuzzy set membership but differs from all of them. The global typicality, on the other hand, resembles well-known histograms but also differs from them. A distinctive feature of the proposed new approach, empirical data analysis (EDA), is that it is not limited by restrictive impractical prior assumptions about the data generation model as the traditional probability theory and statistical learning approaches are. Moreover, it does not require an explicit and binary assumption of either randomness or determinism of the empirically observed data, their independence, or even their number (it can be as low as a couple of data samples). The typicality is considered as a fundamental quantity in the pattern analysis, which is derived directly from data and is stated in a discrete form in contrast to the traditional approach where a continuous pdf is assumed a priori and estimated from data afterward. The typicality introduced in this paper is free from the paradoxes of the pdf. Typicality is objectivist while the fuzzy sets and the belief-based branch of the probability theory are subjectivist. The local typicality is expressed in a closed analytical form and can be calculated recursively, thus, computationally very efficiently. The other nonparametric ensemble properties of the data introduced and studied in this paper, namely, the square centrality, cumulative proximity, and eccentricity, can also be updated recursively for various types of distance metrics. Finally, a new type of classifier called naïve typicality-based EDA class is introduced, which is based on the newly introduced global typicality. This is only one of the wide range of possible applications of EDA including but not limited for anomaly detection, clustering, classification, control, prediction, control, rare events analysis, etc., which will be the subject of further research.

The Yellow River basin is a key food production centre of global importance facing rapidly growing water scarcity. Water availability for agriculture in the basin is threatened by rapid growth in the demand for industrial and urban water, the need to flush sediment from the river's lower reaches, environmental demands and growing water pollution. Climate change is already evident in the basin with long-term declines in river runoff, higher temperatures, and increasing frequency and intensity of drought. The Chinese government has exhausted most options for improving water supply. The challenge will be to switch to improved water demand management, which is hampered by existing governance structures, and lack of integrated agriculture and water resource policies.

With the development of power electronics, calculator technique and the micro-control technology, motor has been applied widely in various fields. High-performance motor control requires rotor speed to implement field-oriented control. These motor speed measurement methods have their own advantages and disadvantages. A method based on second order generalized integrator-frequency locked loop (SOGI-FLL) is proposed in this paper. It is consist of second order generalized integrator quadrature signal generator (SOGI-QSG) and frequency locked loop (FLL). The control principle of this algorithm is analyzed and the model was established by simulink. The analysis results show that this algorithm calculates the frequency of input signals very quickly. When the input signal is triangle wave, this algorithm is still applicable. It has good performance of frequency estimation.

Pseudospark discharges are sources of intense electron beams. Reported in this paper are theoretical studies of the pseudospark discharges. We have developed a theoretical and computational model to study the initiation phase of pseudospark discharges via PIC-MCC method. From the numerical results,we have identified a sequence of physical events. In spite of the approximation in this model, the picture which has emerged provides insight into the virtual anode moving which were observed by previous experiments.