Search

Depression Detection of Speech is widely applied due to its ease of acquisition and imbuing with emotion. However, there exist challenges in effectively segmenting and integrating depressed speech segments. Multiple merges can also lead to blurred original information. These problems diminish the effectiveness of existing models. This article proposes a Hierarchical Transformer model for speech depression detection based on dynamic window and attention merge, abbreviated as DWAM-Former. DWAM-Former utilizes a Learnable Speech Split module (LSSM) to effectively separate the phonemes and words within an entire speech segment. Moreover, the Adaptive Attention Merge module (AAM) is introduced to generate representative feature representations for each phoneme and word in the sentence. DWAM-Former also associates the original feature information with the merged features through a Variable-Length Residual module (VL-RM), reducing feature loss caused by multiple mergers. DWAM-Former has achieved highly competitive results in the depression detection dataset DAIC-WOZ. An MF1 score of 0.788 is received in the experiment, representing a 7.5% improvement over previous research. [ABSTRACT FROM AUTHOR]

The advent of the peer-to-peer energy trading has dramatically altered the conventional dynamics of distribution networks. Prosumers, managing independently and engaging in peer-to-peer (P2P) energy exchanges, introduce critical challenges for the economical and secure operations of these networks. This research presents a hierarchical framework designed to manage P2P energy interactions between prosumers and to facilitate flexible operation within the distribution network at the substation level. The upper layer of the model aims to stabilize market prices within the distribution framework, while the lower layer establishes a P2P energy trading platform that upholds fairness and safeguards the privacy of the prosumer participants. A fixed-point mapping approach is utilized to assess the interactions between market stabilization efforts and prosumer bidding strategies within this framework. Through simulations, we illustrate the logical soundness and effectiveness of our proposed model and approach. The findings indicate that the proposed model and the energy trading framework could substantially improve the overall welfare of all stakeholders involved. [ABSTRACT FROM AUTHOR]

Paced mixed-model assembly lines are popular with various manufacturing enterprises. However, they face the weakness that they are at risk of line stoppages owing to the occurrence of workstation work overload situations. Moreover, the consideration of workers' health and ergonomic risk on manual assembly lines is a necessity required by legislation. Therefore, this article addresses mixed-model assembly line balancing and sequencing taking the problem of ergonomic risk into consideration and manages work overloads using a side-by-side policy with utility workers. To solve these problems, this article proposes an hierarchical and two-stage framework. The optimization goals are to minimize the number of workstations and utility workers. Furthermore, this article integrates an iterated greedy algorithm into a genetics algorithm to obtain global exploration and local exploitation ability. Finally, a divide-and-conquer strategy is proposed to meet the challenge of solving a large-scale problem. Experimental results show the effectiveness of the mechanism proposed in this article. [ABSTRACT FROM AUTHOR]

Depression Detection of Speech is widely applied due to its ease of acquisition and imbuing with emotion. However, there exist challenges in effectively segmenting and integrating depressed speech segments. Multiple merges can also lead to blurred original information. These problems diminish the effectiveness of existing models. This article proposes a Hierarchical Transformer model for speech depression detection based on dynamic window and attention merge, abbreviated as DWAM-Former. DWAM-Former utilizes a Learnable Speech Split module (LSSM) to effectively separate the phonemes and words within an entire speech segment. Moreover, the Adaptive Attention Merge module (AAM) is introduced to generate representative feature representations for each phoneme and word in the sentence. DWAM-Former also associates the original feature information with the merged features through a Variable-Length Residual module (VL-RM), reducing feature loss caused by multiple mergers. DWAM-Former has achieved highly competitive results in the depression detection dataset DAIC-WOZ. An MF1 score of 0.788 is received in the experiment, representing a 7.5% improvement over previous research.

Abstractive summarization is one of the popular topics that has been the researchers' attention for several years. This is because of the widespread application frameworks included in this field. Most of the existing summarization frameworks cannot provide effective abstracts as the contextual information of the input is not given importance. To deal with the problem, this work introduces a hierarchical framework using transformer technology to produce effective abstracts. The proposed framework includes preprocessing, extractive summarization, and abstractive summarization as the basic steps of the work. Initially, the input contents are preprocessed to obtain a clean document, and then the contents are provided to the extractive summarization unit. This unit consists of a fine-tuned BERTSum model (FTBS), which is a pre-trained model to produce the required extractive summary. The output is then provided to the proposed convolutional bidirectional gated recurrent unit transformer (CBi-GRUT) model, where an additional encoder model is introduced with the traditional transformer technology to obtain the output. The outcomes of the model are then assessed with the existing models to prove its efficacy, and the evaluations are carried out using the CNN/Daily Mail dataset. The proposed method achieved an average ROUGE-1 score of 0.78, average ROUGE-2 score of 0.68 and an average ROUGE-L score of 0.77. [ABSTRACT FROM AUTHOR]

Great progress has been made in the domain of multi-agent reinforcement learning in recent years. Most work concentrates on solving a single task by learning the cooperative behaviors of agents. However, many real-world problems are normally composed of a set of subtasks in which the execution order follows a certain procedure. Cooperative behaviors should be learned on the premise that agents are first allocated to those subtasks. In this paper, we propose a hierarchical framework for learning the dynamic allocation of agents among subtasks, as well as cooperative behaviors. We design networks corresponding to agents and subnetworks, respectively, which constitute the whole hierarchical network. For the upper layer, a novel allocation learning mechanism is devised to map an agent network to a subtask network. Each agent network could only be assigned to only one subtask network at each time step. For the lower layer, an action learning module is designed to compute appropriate actions for each agent with the allocation result. The agent networks together with the subtask networks are updated by a common reward obtained from the environment. To evaluate the effectiveness of our framework, we conduct experiments in two challenging environments, i.e., Google Research Football and SAVETHECITY. Empirical results show that our framework achieves much better performance than other recent methods. [ABSTRACT FROM AUTHOR]

Abstract Great progress has been made in the domain of multi-agent reinforcement learning in recent years. Most work concentrates on solving a single task by learning the cooperative behaviors of agents. However, many real-world problems are normally composed of a set of subtasks in which the execution order follows a certain procedure. Cooperative behaviors should be learned on the premise that agents are first allocated to those subtasks. In this paper, we propose a hierarchical framework for learning the dynamic allocation of agents among subtasks, as well as cooperative behaviors. We design networks corresponding to agents and subnetworks, respectively, which constitute the whole hierarchical network. For the upper layer, a novel allocation learning mechanism is devised to map an agent network to a subtask network. Each agent network could only be assigned to only one subtask network at each time step. For the lower layer, an action learning module is designed to compute appropriate actions for each agent with the allocation result. The agent networks together with the subtask networks are updated by a common reward obtained from the environment. To evaluate the effectiveness of our framework, we conduct experiments in two challenging environments, i.e., Google Research Football and SAVETHECITY. Empirical results show that our framework achieves much better performance than other recent methods.

The sparse reward problem has long been one of the most challenging topics in the application of reinforcement learning (RL), especially in complex multi-agent systems. In this paper, a hierarchical multi-agent RL architecture is developed to address the sparse reward problem of cooperative tasks in continuous domain. The proposed architecture is divided into two levels: the higher-level meta-agent implements state transitions on a larger time scale to alleviate the sparse reward problem, which receives global observation as spatial information and formulates sub-goals for the lower-level agents; the lower-level agent receives local observation and sub-goal and completes the cooperative tasks. In addition, to improve the stability of the higher-level policy, a channel is built to transmit the lower-level policy to the meta-agent as temporal information, and then a two-stream structure is adopted in the actor-critic networks of the meta-agent to process spatial and temporal information. Simulation experiments on different tasks demonstrate that the proposed algorithm effectively alleviates the sparse reward problem, so as to learn desired cooperative policies. [ABSTRACT FROM AUTHOR]

Wireless Sensor Networks (WSNs) and the Internet of Things (IoT) have emerged as transforming technologies, bringing the potential to revolutionize a wide range of industries such as environmental monitoring, agriculture, manufacturing, smart health, home automation, wildlife monitoring, and surveillance. Population expansion, changes in the climate, and resource constraints all offer problems to modern IoT applications. To solve these issues, the integration of Wireless Sensor Networks (WSNs) and the Internet of Things (IoT) has come forth as a game-changing solution. For example, in agricultural environment, IoT-based WSN has been utilized to monitor yield conditions and automate agriculture precision through different sensors. These sensors are used in agriculture environments to boost productivity through intelligent agricultural decisions and to collect data on crop health, soil moisture, temperature monitoring, and irrigation. However, sensors have finite and non-rechargeable batteries, and memory capabilities, which might have a negative impact on network performance. When a network is distributed over a vast area, the performance of WSN-assisted IoT suffers. As a result, building a stable and energy-efficient routing infrastructure is quite challenging in order to extend network lifetime. To address energy-related issues in scalable WSN-IoT environments for future IoT applications, this research proposes EEDC: An Energy Efficient Data Communication scheme by utilizing "Region based Hierarchical Clustering for Efficient Routing (RHCER)"—a multi-tier clustering framework for energy-aware routing decisions. The sensors deployed for IoT application data collection acquire important data and select cluster heads based on a multi-criteria decision function. Further, to ensure efficient long-distance communication along with even load distribution across all network nodes, a subdivision technique was employed in each tier of the proposed framework. The proposed routing protocol aims to provide network load balancing and convert communicating over long distances into shortened multi-hop distance communications, hence enhancing network lifetime.The performance of EEDC is compared to that of some existing energy-efficient protocols for various parameters. The simulation results show that the suggested methodology reduces energy usage by almost 31% in sensor nodes and provides almost 38% improved packet drop ratio. [ABSTRACT FROM AUTHOR]

Purpose: The research aims to explore the wisdom, knowledge and practices in vernacular housing settlements with their sustainability underpinnings as tools for modelling rural affordable housing in tropical regions. The study is based on a weaving settlement in Bargarh district of Odisha, which is globally acclaimed for its Ikkat style of weaving. Design/methodology/approach: A hierarchical framework of sustainability resting on the three pillars of ecological, economical and environmental dimensions is derived from existing theoretical research. This framework of 22 indicators is subsequently assigned to assess the sustainability of the vernacular weavers' settlement through quantitative evaluation. A qualitative assessment through observation and deduction also verifies the result. Findings: Since the vernacular weavers settlement performs very well on the sustainability scorecard, the paper suggests that its best practices can be incorporated while designing affordable housing so that social, cultural and heritage values are retained and a climate conscious, energy-efficient sustainable approach is ensured. Practical implications: The recommendations from the assessment has many lessons while framing policies for rural affordable housing as it cannot have one size that fits all settlement typology irrespective of the occupational, climatic and social needs. Originality/value: The sustainable design and planning principles embedded in this vernacular settlement offers a valuable blueprint to re-imagine the affordable housing in rural areas which can be myopic if it does not take into account the occupational needs and life style of craftsmen dwellers. [ABSTRACT FROM AUTHOR]

This study contributes to the complex adaptive system theory by offering a valid hierarchical model to evaluate the theory's important features related to resilience. The garment industry in Bangladesh encountered disruption in the supply chain during the COVID-19 pandemic and the supply chain competencies played a vital role in overcoming the crisis. Limited studies are built on a solid theoretical foundation and considered supply chain competencies in assessing supply chain resilience. This study aims to develop a multi-criteria hierarchical measurement structure by considering the supply chain competencies to evaluate supply chain resilience. Fuzzy Delphi method and Fuzzy importance and performance analysis approach were applied for the study purpose. Findings reveal health and safety management, information management system, business intelligence, innovation capabilities management, technological innovation, and artificial intelligence as critical criteria, and data, information, and computing, technological innovation and adaptation are critical aspects that require improvement. [ABSTRACT FROM AUTHOR]

The advent of same-day delivery services has achieved immense popularity, driven by escalating customer expectations on fast shipping and the need for market competitiveness. To optimize such services, the use of heterogeneous fleets with vehicles and drones has proven effective in reducing the resource requirements needed for delivery. This paper focuses on investigating the same-day delivery dispatching and routing problem with a fleet of multiple vehicles and drones. In this problem, stochastic and dynamic requests, coupled with their stringent time constraints, require dispatchers to make real-time decisions about optimally assigning vehicles and drones, ensuring both efficiency and effectiveness in delivery operations while taking into account the routing. To tackle this complex problem, we model it with a route-based Markov decision process and develop a novel hierarchical decision approach based on deep reinforcement learning (HDDRL). The first level of the hierarchy is tasked with determining the departure times of vehicles, balancing the trade-offs between the delivery frequency and efficiency. The second level of the hierarchy is dedicated to determining the most suitable delivery mode for each request, whether by vehicles or drones. The third level is responsible for planning routes for vehicles and drones, thereby enhancing route efficiency. These three levels in the hierarchical framework collaborate to solve the problem in a synchronized manner, with the objective of maximizing the service requests within a day. Empirical results from computational experiments highlight the superiority of the HDDRL over benchmark methods, demonstrating not only its enhanced efficacy but also its robust generalization across diverse data distributions and fleet sizes. This underscores the HDDRL's potential as a powerful tool for enhancing operational efficiency in same-day delivery services. • A new problem for dispatching and routing in same-day delivery with vehicles and drones. • Formulate the same-day delivery problem with a route-based MDP. • A novel hierarchical decision approach based on deep reinforcement learning. • Strong generalization ability across different data distribution and fleet sizes. [ABSTRACT FROM AUTHOR]

The increasing adoption of electric vehicles (EVs) poses significant challenges for power system operations, requiring scalable coordination to mitigate their negative impacts and leverage their potential to enhance grid conditions. This paper introduces a scalable, three-layer hierarchical framework for optimal EV charge and discharge scheduling (EVCDS) that coordinates key agents: EVs, EV aggregators (EVAs), and the distribution network operator (DNO). The optimization problem is developed as an exchange problem and solved using the alternating direction method of multipliers (ADMM) in a decentralized approach. The proposed EVCDS addresses economic factors by minimizing battery degradation costs at the EV level and charging costs at the EVA level, while managing technical aspects at the DNO level by minimizing load curve variance and limiting power capacity. Moreover,voltages at network nodes are calculated using the DistFlow model to simplify the optimization and ensure compliance with standard operational limits. Compared to uncoordinated EV charging, EVCDS reduces load profile deviations by 85% and total costs by 91%, while also improving bus voltage profiles. [ABSTRACT FROM AUTHOR]

Many methods have been proposed to extract line segment (LS) correspondences for images with viewpoint variations. However, the matching performance on images with viewpoint variations is still limited due to the influence of discontinuous parallax and poor textures on images. Existing line segment matching (LSM) methods rarely consider the coexistence of these two situations. This study attempts to address this problem by proposing an adaptive feature region (FR)-based LSM method based on the observation that different FRs are suitable for different image scenes under image viewpoint change. In the proposed method, LSs are paired, and two types of FRs, symmetric FR centered on the line-pair intersection and asymmetric FR with the intersection as a vertex, are constructed for every line pair. The asymmetric FR of a line pair is a subregion of its symmetric FR. And the complement region of asymmetric FR in symmetric FR is called complement FR. To choose a suitable FR for the matching of each line pair adaptively, the asymmetric FR-based feature descriptor similarity and epipolar line-based constraints are combined to determine candidate line pair matches firstly. Then, the candidate matches with similar complement FR-based feature descriptors are considered initial matches and used to construct a topological constraint to check other candidates. Finally, we estimate an adaptive influence region-based local homography for each matched line pair to constrain the matching of unmatched individual LSs. The experimental results on the test images with viewpoint changes show that our approach outperforms the state-of-the-art methods in terms of Recall, matching precision (MP), and F-Measure (reflecting the overall matching performance of Recall and MP). Especially, it improves the average Recall and average F-Measure of the best method among the comparison methods by 45.20% and 25.83%, respectively.

Gait plays a decisive role in the performance of hexapod robot walking; this paper focuses on adaptive gait generation with reinforcement learning for a hexapod robot. Moreover, the hexapod robot has a high-dimensional action space and therefore it is a great challenge to use reinforcement learning to directly train the robot's joint angles. As a result, a hierarchical and modular framework and learning details are proposed in this paper, using only seven-dimensional vectors to denote the agent actions. In addition, we conduct experiments and deploy the proposed framework using a real hexapod robot. The experimental results show that superior reinforcement learning algorithms can converge in our framework, such as SAC, PPO, DDPG and TD3. Specifically, the gait policy trained in our framework can generate new adaptive hexapod gait on flat terrain, which is stable and has lower transportation cost than rhythmic gaits. [ABSTRACT FROM AUTHOR]

Satellite attitude determination, as a crucial pre-processing technology in remote sensing earth observation, is significantly associated with the geometric accuracy of satellite stereo images. To compensate for the observation errors in raw attitude data for achieving accurate results, the conventional Kalman filter-based and filter-smoothing combination frameworks provide insufficient access to statistical features of data, resulting in inferior practical system performance. To conquer these drawbacks, a hierarchical and efficient attitude determination framework is proposed called HE 2 LM-AD, which consists of a simplified adaptive Kalman filtering module, a neural network-based system error compensation module, and a weighted attitude smoothing module. To be more specific, an adaptive noise covariance matrix update is exploited in the Extended Kalman Filter (EKF) iterations, and a simplification scheme based on Cholesky decomposition is employed to obtain coarse filter outputs efficiently. Following that, a dedicated Extreme Learning Machine (ELM)-based error compensation network is designed to capture the pattern of observation errors, providing robust compensation at lower overhead. To further pursue high-performance attitude determination results, we fused the outcomes of the aforementioned two components according to the weights achieved from adaptive learning. Experiment results on in-orbit satellite attitude simulation datasets demonstrate and validate the feasibility and effectiveness of our framework. [ABSTRACT FROM AUTHOR]

This study attempts to identify critical challenges for blockchain adoption in government, particularly public-service delivery in India, a developing country context. Through an extensive literature review and focus-group discussions with policymakers and blockchain experts, we have identified 12 adoption challenges for Blockchain in public service delivery. We then collected data and analysed using interpretive structural modeling and Cross-Impact Matrix Multiplication Applied to Classification (MICMAC) Analysis to develop a hierarchical framework of the challenges. Our findings indicate that governments must first ensure legislative support for blockchain-based transactions. This research contributes to information systems strategic planning literature and provides a framework for policymakers to craft a strategic approach to facilitate blockchain adoption. [ABSTRACT FROM AUTHOR]

Broad learning system (BLS) is an emerging machine learning algorithm with high efficiency and good approximation capability. It has been proved that BLS can learn hundreds of times faster than traditional deep learning algorithms while providing a comparable or even better generalization performance. Owing to its superb efficiency and powerful learning ability, the BLS is attracting increasing attention from machine learning community and can be considered as an alternative to deep learning in some situations. However, due to its shallow structure, the feature learning of BLS is not sufficient and which may probably limit its learning performance. For this issue, this paper proposes a novel hierarchical broad learning system (H-BLS) with deep and sparse feature learning. Different from the original BLS which conducts feature learning simply using a single-layer function mapping, the H-BLS adopts a hierarchical feature learning framework with multi-layer and multi-group structure to extract high-level and rich feature information from the original input, so as to improve the feature representation capability of the model. Meanwhile, in the hierarchical feature learning process of H-BLS, a new l1-constrained sparse autoencoder is employed and embedded in each layer of the framework for feature reconstruction, so as to eliminate redundancy of the input and generate more sparse and compact feature representations, thus further enhancing its learning performance. The learning ability of the proposed H-BLS is firstly evaluated by ten commonly used regression data sets, and the experimental results show that H-BLS performs better compared with several representative learning algorithms such as SVM, LSSVM, ELM, BLS and two recently proposed BLS variants. Moreover, the H-BLS also shows advantages over the state-of-the-art methods in terms of classification accuracy and training time on image classification problems. [ABSTRACT FROM AUTHOR]

Recent work on intelligent agents is a popular topic among the artificial intelligence community and robotic system design. The complexity of designing a framework as a guide for intelligent agents in an unknown built environment suggests a pressing need for the development of autonomous agents. However, most of the existing intelligent mobile agent design focus on the achievement of agent's specific practicality and ignore the systematic integration. Furthermore, there are only few studies focus on how the agent can utilize the information collected in unknown build environment to produce a learning pipeline for fundamental task prototype. The hierarchical framework is a combination of different individual modules that support a type of functionality by applying algorithms and each module is sequentially connected as a prerequisite for the next module. The proposed framework proved the effectiveness of ESNI system integration in the experiment section by evaluating the results in the testing environment. By a series of comparative simulations, the agent can quickly build the knowledge representation of the unknown environment, plan the actions accordingly, and perform some basic tasks sequentially. In addition, we discussed some common failures and limitations of the proposed framework. [ABSTRACT FROM AUTHOR]

A holistic fault detection and diagnostics (FDD) method should explicitly consider the dependencies between faults at the system- and zone-level to isolate the root cause. A system-level fault can trigger false alarms at the zone-level, while concealing the presence of a zone-level fault. However, most FDD methods have focused on a single component/equipment without considering the importance of the interactions between zone- and system-level devices. This paper proposes a holistic hierarchical framework for FDD, combining the process of detection and diagnosis of controls hardware and sequencing logic faults affecting the actuators at the system- and zone-level. The proposed framework follows a holistic sequential procedure to diagnose faults and suppress false alarms in this order: hard faults in air handling units (AHUs), hard faults in variable air volume (VAV) zones, sequencing logic faults in AHUs, and sequencing logic faults in VAV zones. The detection of faults is performed by visualizing the discrepancies between the expected and measured operational behaviour of AHUs and VAV boxes. Examples demonstrating the framework are provided with data from 10 different VAV AHU systems. Practical application: This paper provides a sequential hierarchical FDD framework to address two main issues in VAV AHU systems: detectability and significance. Regarding detectability, the framework prioritizes hard faults over sequencing logic faults to avoid false positives and false negatives; about significance, system-level faults are prioritized over zone-level faults to triage high-impact faults in the system. The detection of faults is performed via visualizing the biases from the expected behaviour of AHU and VAV characteristics to provide an envisioning interpretation for the experts in facilities management in commercial buildings to find the root cause of the fault and fix them on-site. [ABSTRACT FROM AUTHOR]

It is important to understand the ultimate control of COVID-19 in all countries around the world in relation to the characteristics of developed countries, LDCs, and the variety of transmission characteristics of COVID-19. Therefore, this study aimed to identify factors associated with confirmed cases of COVID-19 with a focus on the Human Development Index (HDI). The units of analysis used for the current study were countries, and dataset were aggregated from multiple sources. This study used COVID-19 data from Our World in Data, the Global Health Security Index, and the WORLD BANK. A total of 171 countries were included in the analysis. A multi-variable linear regression with a hierarchical framework was employed to investigate whether the HDI is associated with confirmed COVID-19 cases after controlling for the demographic and healthcare system characteristics of the study countries. For Model 2, which controlled for demographic and healthcare system characteristics, HDI (β = 0.46, p < 0.001, 95% CI = 2.64–10.87) and the number of physicians per 1000 people (β = 0.34, p < 0.01, 95% CI = 0.21–0.75) had significant associations with the total number of confirmed COVID-19 cases per million people. Countries with a high HDI level are able to conduct higher per capita testing, resulting in higher numbers of confirmed cases than in countries with lower HDI levels. This study has shown evidence that could be used by governments and international organizations to identify national characteristics and provide the international cooperation necessary to develop effective prevention and intervention methods to deal with the global pandemic. [ABSTRACT FROM AUTHOR]

In the real-time status monitoring of pipeline network, incomplete pressure data are unavoidable due to some device or communication errors. To solve this problem, a hierarchical data recovery method based on generative adversarial networks (GANs) is proposed in this article. First, a hierarchical data recovery framework is proposed to handle different numbers of incomplete data due to the structure of the semicentral pipeline network. Second, a joint attention module is presented to capture both interior nature and correlation relationships of multivariate pressure series and further guarantee the consistency of pressure data. Third, the macromicrodual discriminators are proposed to evaluate the recovery result through the combination of the local and global variation in temporal and spatial dependencies. Based on the novel structures, the proposed model is able to recover incomplete data with abnormal fluctuation values, unreasonable fixed values, or missing values. Finally, under a series of data recovery experiments, the efficiency of the proposed method is evaluated. Experimental results demonstrate that the proposed method is a practical way to ensure data recovery performance in the pipeline network. Note to Practitioners—Status monitoring based on pressure data is of great importance for safe and efficient operation in a pipeline network. However, due to unexpected situations, the appearance of incomplete pressure data affects the subsequent data processing and status analysis, resulting in an incorrect decision. In this article, a deep learning-based method is proposed to recover the incomplete data. With the help of the spatiotemporal dependencies of multivariate pressure series, the proposed method can recover different numbers of incomplete data through the no-missing part of pressure data. The experiment results show that the proposed method is better than the similar data recovery methods through three different evaluation metrics. In the future, we will address the data recovery problem without the complete data pairs in the training process. [ABSTRACT FROM AUTHOR]

The view of Working Memory (WM) as a conscious process has allowed defining consciousness as the content of working memory. However, concerns have emerged over comparisons between consciousness and working memory. Goal: although the relationship between these two study fields has been the matter of psychology, philosophy and neuroscience, a theoretical review addressing the core elements of highly cited perspectives would enrich the discussion in this study area. Method: this review focuses on three theoretical frameworks: 1) the multi-component model of working memory, 2) the global workspace theory, 3) the hierarchical framework. The authors analyzed 113 articles which discussed the previous three models. Results: the multi-component model of working memory contributes a basic functional description on how mental representations remain on-line during complex cognitive processing. Thereby, the information exchange between the central executive and the episodic buffer, in one sense, and the phonological loop and the visuo-spatial sketchpad in the other is given through conscious processing. Conclusions: likewise, the central executive controls and changes attention but the episodic buffer allows multimodal information availability. [ABSTRACT FROM AUTHOR]

In recent years, Smart Cities and Smart Homes have been studied as an important field of research. The design and construction of smart homes have flourished so that this technology is used for more comfort of people and helping their health. This technology is very useful for single and elderly people especially for Alzheimer's disease patients who are alone at home and need permanent care. A lot of research has been done to detect the activity of users in smart homes. They gather raw data of sensors and use the usual classification algorithms for activity detection. The essence of the time dependency of sensors' data has been ignored in most research, while considering this is very important, especially in the case of Alzheimer's patients. In this study, a Nonlinear AutoRegressive Network (NARX) is employed to detect the patient's activity in a smart home. NARX is a recurrent dynamic network, which is commonly used in time-series modeling. The results show that the proposed model detects user activity, with an accuracy of 0.98. Since the high-risk behavior of the Alzheimer's patient is very unknown; a fuzzy inference system is implemented based on the experience of the Alzheimer's sub-specialist and nurses. The main parameters were extracted and a 3-layers hierarchical fuzzy inference system was developed to detect and alarm the patient's high-risk behavior without wearable sensors. The results show 98% accuracy in detecting the patient's activity and 84% accuracy in determining its abnormality. [ABSTRACT FROM AUTHOR]

Contexts: It has been widely acknowledged that exposure to green space (e-GS) has positive health benefits to urban residents. While most studies estimate e-GS from an availability or accessibility perspective, few studies have considered GS quality, which is closely related to the willingness and time of residents' visits to GS. Objectives: Here we propose a hierarchical framework to assess residential e-GS including three individual indicators—availability, accessibility, and attractiveness, and further explore the disparities in GS exposure across rural, peri-urban, and urban areas in a rapidly urbanizing Chinese city (Yangzhou). Specifically, availability means assessing the quantity of surrounding greenness including all types of GSs; accessibility means calculating the network distance from home to major GS with recreational facilities (e.g., public parks); and attractiveness indicator integrated the major GS 'micro' features (i.e., quality), proximity and population density calculated by a modified gravity model. Results: The results show the spatial distribution of residential e-GS was different among availability, accessibility, and attractiveness, and these metrics showed weak correlations suggesting they are three distinct e-GS metrics. Significant differences (Kruskal–Wallis test, p < 0.01) were revealed in the comparisons of the GS availability, accessibility, and attractiveness values among the urban, peri-urban, and rural areas. Further, we outlined the potentially preferable exposure metrics in exploring the pathways linking GS to various dimensions of health outcomes. Conclusions: The hierarchical framework has important theoretical and practical significance in identifying the hierarchical form of e-GS and targeting vulnerable communities that may suffer from health issues due to lack of GSs. [ABSTRACT FROM AUTHOR]

Developing high-efficiency and low-cost catalysts towards oxygen evolution reaction (OER) is extremely important for overall water splitting and rechargeable metal−air batteries. Herein we propose a promising organometallic coordination polymer (OCP) induced strategy to construct hierarchical N-doped carbon framework with NiFe nanoparticles encapsulated inside (NxFe@N–C) as a highly active and stable OER catalyst. The synthesis of OCP precursor depends on the unique molecular structure of iminodiacetonitrile (IDAN), which can coordinate with metal ions to form Ni2Fe(CN)6 with prussian blue analogs (PBA) structure. Unlike previous PBA-induced methods, the thickness of the carbon layer covering the surface of the metal core can be well controlled during the pyrolysis through adjusting the amount of IDAN, which builds a wonderful bridge for investigating the relationship between carbon layer thickness and catalytic performance. Both the experimental characterizations and theoretical studies validate that a suitable carbon layers thickness leads to optimal OER activity and stability. By optimizing the structure and composition, the optimized Ni3Fe@N–C with hierarchical framework exhibits the low overpotentials (260 ​mV at 10 ​mA ​cm−2; 320 ​mV at 50 ​mA ​cm−2), improved kinetics (79 ​mV dec−1), and robust long-term stability, which exceeds those of benchmark RuO2.

Gait plays a decisive role in the performance of hexapod robot walking; this paper focuses on adaptive gait generation with reinforcement learning for a hexapod robot. Moreover, the hexapod robot has a high-dimensional action space and therefore it is a great challenge to use reinforcement learning to directly train the robot’s joint angles. As a result, a hierarchical and modular framework and learning details are proposed in this paper, using only seven-dimensional vectors to denote the agent actions. In addition, we conduct experiments and deploy the proposed framework using a real hexapod robot. The experimental results show that superior reinforcement learning algorithms can converge in our framework, such as SAC, PPO, DDPG and TD3. Specifically, the gait policy trained in our framework can generate new adaptive hexapod gait on flat terrain, which is stable and has lower transportation cost than rhythmic gaits.

Line segment matching for wide-baseline images is challenging due to the significant viewpoint differences. In this study, we propose a hierarchical line segment matching method based on viewpoint robust local structure and geometric constraints. In our approach, line segments are paired and classified into three types representing heuristically those with different level of matchability: structured line pairs (S-LPs), unstructured line pairs (U-LPs), and individual line segments (I-LSs). Accordingly, we design a hierarchical matching framework that consists of three matching layers respectively corresponding to the above three types: in the first layer, robust local structures are constructed for S-LPs. We match the S-LPs by measuring local structure similarity (LSS). In the second layer, we build a topological descriptor-based constraint based on the S-LP matches and combine it with epipolar geometry-based constraints to select candidate matches for U-LPs, and then use LSS to further refine the matches. In the third layer, we estimate local homography for I-LSs to build constraints, to extract matches by exploiting the implicit region information of line pair matches. A pair of pre- and postprocessing algorithms, namely line segment merging and match reassignment, are performed before and after the matching procedure to overcome the negative effect of line segment fragmentation on the matching. Experimental results demonstrate that the proposed method performs better than state-of-the-art methods (with the largest improvement of 124.17% in terms of F-Measure over the best one among the compared methods). [ABSTRACT FROM AUTHOR]

Biped wall-climbing robots well adapt for carrying out tasks in large-scale 3-D environments such as skeleton-facade structures, due to their good abilities to transit between walls and surmount obstacles. However, to plan 3-D collision-free climbing paths in such environments is challenging and significant for biped wall-climbing robots. In this article, we propose a novel hierarchical framework for planning 3-D biped climbing navigation, including three levels, namely the global route, the foothold, and the single-step motion. We also present three effective planners incorporated in the framework to solve this challenging planning problem efficiently. A global route planner is designed on top, for rapidly analyzing the possibility for the robot to transit between walls, and searching walls that must be passed through and key footholds for performing transition, in order to optimize the route length globally. A local foothold planner is implemented in the middle, to search and optimize footholds on each via wall, considering the robot's ability to step over, step on, and bypass obstacles. A single-step motion planner is deployed at the bottom, for planning collision-free movement swinging the robot suction module from a foothold to another. Simulations are conducted to fully verify the effectiveness and performance of the presented planners. Experiments with W-Climbot are carried out to demonstrate that the proposed framework and planners succeed in planning 3-D collision-free optimized climbing paths for biped wall-climbing robots. [ABSTRACT FROM AUTHOR]

Learning to perform air combat autonomously has been a long-standing challenge. The design of intelligent game strategies was difficult due to complex dynamic constraints and long decision-making process. Most existing approaches depend heavily on simplified aircraft models or complex hand-crafted rules. Different from previous works, this paper presents H3E, a novel and efficient Three-level Hierarchical decision framework embedding Expert knowledge, which gives birth to various strategies for high-fidelity one-on-one beyond-visual-range (BVR) air combat game. Inspired by the way pilots make decisions, we build a hierarchical framework to divide the air combat into several sub systems in which each level can perform its own task with much smaller exploration space. In addition, to make full use of the expert knowledge while minimizing its limitation, a novel "Rule-Imitation-Reinforcement" (RIR) training paradigm with an adjustable expert-guidance (AEG) loss function is established, which can increase the exploration efficiency as well as the game win rate. Finally, this work is evaluated in the Intelligent Air Game Simulator (IAGSim), a high-fidelity air combat simulation platform, through a series of games against the state-of-the-art (SOTA) methods. The learning process and game results verify the superior performance of our framework in terms of the exploration efficiency (higher rewards with the same training samples) and win rate (at least 72.9%) compared with the existing SOTA approaches. • A novel three-level hierarchical decision framework for air combat is proposed. • A 'Rule-Imitation-Reinforcement' training paradigm is newly introduced. • Making full use of expert knowledge via an adjustable expert-guidance loss. • Experiments show better exploration efficiency and game performance. • Interpretable strategies have emerged under the training framework. [ABSTRACT FROM AUTHOR]

The Internet of Things establishes the intimacy between the Internet and the physical world. Due to portable size, most IoT devices have limited computing and storage capabilities and are vulnerable to various malicious intrusions. Therefore, it is vital to have efficient approaches to distinguish the true IoT data from fake one, we term such methods as anomaly detection (AD). To detect anomalies accurately and efficiently, in this article a 3-hierarchy joint local and global anomaly detection framework, HADIoT, is proposed, in which IoT devices generate and transmit sensory data to their local edge servers for local AD after data refinement which includes re-framing, normalization, complexity reduction via Principal Component Analysis, and symbol mapping. High detection accuracy is achieved by jointly local and global ADs. The local AD focuses on the data pattern consistency of individual devices via the Gated Recurrent Unit, and the processed data is then forwarded from edge servers to the cloud server for global AD. The global AD focuses on the analysis of the data pattern correlations between different IoT devices, using the Conditional Random Fields. For the maintenance of cyber-security, the proposed anomaly detection framework HADIoT enables to provide an accurate and faster anomaly detection for IoT applications, compared with existing anomaly detection methods. The performance of the proposed method is also empirically evaluated through simulations, using a real dataset - the Information Security Center of Excellence (ISCX) 2012 dataset. Simulation results demonstrate the effectiveness of the proposed framework in terms of True Positive Rate, False Positive Rate, Precision, Accuracy and F_score, compared with three benchmark schemes.

The emergence of big data technology continues to innovate and change the world, bringing opportunities and challenges to all walks of life. Against the background of this era, traditional entrepreneurship education requires reform and innovation. This research attempts to explore the ways and practices of applying big data technology to entrepreneurship education so as to improve and perfect traditional entrepreneurship education and achieve its sustainable development. Based on classic theories, such as entrepreneurial theory, strategic management theory, and leadership theory, this paper develops a relatively systematic attribute system of entrepreneurship education under big data technology, comprehensively uses Fuzzy-DEMATEL and ISM methods to explore the relationship between different attributes and their importance, and finally constructs a hierarchical framework for the application of big data technology in entrepreneurship education. The results show that the attributes of entrepreneurship education under big data technology can be divided into four levels, each with different priorities and degrees of importance, and there are complex interactions and constraints among them. This study provides important guidance and suggestions for the development of entrepreneurship education and multiattribute decision-making management under the given resources, which is conducive to the sustainable development of entrepreneurs and new ventures.

In recent years, deep reinforcement learning methods have achieved impressive performance in many different fields, including playing games, robotics, and dialogue systems. However, there are still a lot of restrictions here, one of which is the demand for massive amounts of sampled data. In this paper, a hierarchical meta-learning method based on the actor-critic algorithm is proposed for sample efficient learning. This method provides the transferable knowledge that can efficiently train an actor on a new task with a few trials. Specifically, a global basic critic, meta critic, and task specified network are shared within a distribution of tasks and are capable of criticizing any actor trying to solve any specified task. The hierarchical framework is applied to a critic network in the actor-critic algorithm for distilling meta-knowledge above the task level and addressing distinct tasks. The proposed method is evaluated on multiple classic control tasks with reinforcement learning algorithms, including the start-of-the-art meta-learning methods. The experimental results statistically demonstrate that the proposed method achieves state-of-the-art performance and attains better results with more depth of meta critic network.

In actual production practice, the occurrence probability of multiple faults is much higher than that of a single fault. Since the composition of multiple faults is uncertain, it is difficult to establish a single model for multifault diagnosis. In this paper, a new hierarchical framework is proposed for solving multifault detection and isolation problems. First, an adaptive dynamic kernel independent component analysis method is proposed for time-varying and unknown multifault detection. After that, a sparse local exponential discriminant analysis method is developed for the multimodal multifault isolation problem. Finally, the Tennessee Eastman process is used to validate the performance of the proposed methods, and the experimental results show that the proposed methods can efficiently detect and isolate multiple faults.

Many efforts have been focused on the network-wide traffic signal optimization to deal with the congestion problem in big cities. Nevertheless, research evidence illustrates that both improper traffic network managements and excessive traffic demands are the key factors leading to the oversaturated traffic conditions. Current studies encounter the bottleneck in addressing the multi-objective optimization problem. This point calls for designing the hierarchical control framework. In this paper, we concern a two-level hierarchical model-based predictive control scheme to improve mobility in heterogeneous large-scale urban traffic networks, so as to mitigate traffic jams. On the basis of a network partition, a regional demand management approach regulating the input traffic flow from adjacent regions is proposed for multi-subnetworks management taking the advantage of the concept of a macroscopic fundamental diagram of urban traffic networks. This can be viewed as a higher level control layer and can be integrated with other strategies. The lower level control layer utilizes the traffic signals coordination within the subnetworks based on a detailed link-level traffic model to optimize the allocation of vehicles in each subnetwork as homogeneous as possible. The simulation results show that integrating regional demand management with a local traffic responsive control into a hierarchical framework can significantly improve the whole network performance under different traffic scenarios in comparison with other available control strategies.

Multi-access edge computing (MEC) has emerged as a promising paradigm because of its good performance for computation-intensive and latency-critical applications. However, the enormous computing requests from computation service subscribers (CSSs) still cannot be satisfied by pre-existing edge computation nodes (ECNs). To fully utilize the advantage of the MEC network, a hierarchical computation offloading framework is developed under network virtualization (NV) scenario. Accordingly, a two-step sequential process is designed to stimulate the proposed framework. In the first step, an incentive mechanism is proposed in which more temporary ECNs can be motivated by MEC operator and then join the MEC network. Without perfect ECN information, the optimal contract items (the ECN's CPU contribution and reward) between the MEC operator and ECNs can be achieved by taking account of individual rationality (IR) and incentive compatible (IC) constraints. After acquiring the ECNs’ CPU contributions, the computing resource allocation problem between the ECNs and CSSs is then considered in the second step. Since the CSSs have private information, a Bayesian matching game with externality is leveraged to model the problem. Whereas, the conventional resident-oriented Gale-Shapley (RGS) algorithm cannot ensure the stability. Hence, an iterative matching algorithm that can always converge to stable results is developed. Finally, simulation results demonstrate that our proposed two-step sequential decision process can significantly improve social welfare considering the practical scenarios, with reasonable computational complexity. [ABSTRACT FROM AUTHOR]