Article
Computer Science, Artificial Intelligence
Jingliang Duan, Zhengyu Liu, Shengbo Eben Li, Qi Sun, Zhenzhong Jia, Bo Cheng
Summary: This paper presents a constrained adaptive dynamic programming algorithm that can directly handle state-constrained nonlinear nonaffine optimal control problems. By transforming the traditional policy improvement process into a constrained policy optimization problem and approximating the policy and value functions with multi-layer neural networks, the algorithm linearizes the constrained optimization problem and obtains optimal updates.
Article
Automation & Control Systems
Jonghyeok Park, Soo Jeon, Soohee Han
Summary: This article proposes a data-efficient model-based reinforcement learning algorithm that uses reliable future reward estimates achieved through a confidence-based probabilistic ensemble terminal critics. The proposed algorithm chooses actions that optimize the sum of near and distant future rewards, with near future rewards being determined from trained models and distant future rewards assessed using a confidence-based approach. The results show the superiority and data efficiency of the proposed algorithm in both DeepMind Control Suite tasks and real-world control applications.
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS
(2023)
Article
Computer Science, Artificial Intelligence
Zehong Cao, Chin-Teng Lin
Summary: This study investigates the use of global information to accelerate the learning process and increase the cumulative rewards in reinforcement learning (RL) for competition tasks. The proposed RLHC algorithm introduces multiple cooperative critics from a hierarchical framework, allowing agents to access value information from local and global critics. The results show that RLHC outperforms the benchmark algorithm in various competitive tasks.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Computer Science, Information Systems
Dongfen Li, Lichao Meng, Jingjing Li, Ke Lu, Yang Yang
Summary: Deep reinforcement learning has shown excellent performance in robot control, video games, and multi-agent systems. However, most existing models lack generalization capability, limiting their flexibility in real-world applications. To address this issue, this study proposes a two-stage model that focuses on learning adaptation to visual environment changes before optimizing behavioral policies.
INFORMATION SCIENCES
(2022)
Article
Automation & Control Systems
Lei Yan, Zhi Liu, Yun Zhang, Zongze Wu, C. L. Philip Chen
Summary: In this article, an adaptive optimal control scheme is proposed for the strict-feedback nonlinear system, which consists of two design steps. Firstly, a novel nonlinear state-dependent function (NSDF) is formulated to transform the system into a non-constrained one. Secondly, an adaptive optimal control scheme is designed for the non-constrained system using reinforcement learning (RL) to obtain the optimal controller. The effectiveness of the proposed scheme is demonstrated through simulation examples.
Article
Computer Science, Artificial Intelligence
Kaixin Lu, Zhi Liu, Haoyong Yu, C. L. Philip Chen, Yun Zhang
Summary: Solving the problem of optimizing performance and satisfying constraints in control operation usually involves a complicated and time-consuming learning process with neural networks, and is only applicable for simple or time-invariant constraints. This paper proposes a newly adaptive neural inverse approach that removes these restrictions. By introducing a new universal barrier function to transform the constrained system into an equivalent one with no constraint, and designing a switched-type auxiliary controller and a modified criterion for inverse optimal stabilization, it is proven that optimal performance can be achieved with a computationally attractive learning mechanism and all the constraints are never violated. Furthermore, improved transient performance is obtained by explicitly designing the bound of the tracking error. An illustrative example verifies the proposed methods.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Computer Science, Theory & Methods
Dewen Qiao, Songtao Guo, Defang Liu, Saiqin Long, Pengzhan Zhou, Zhetao Li
Summary: This study proposes a distributed resources-efficient proactive content caching (FPC) policy that uses federated learning and deep reinforcement learning. The adaptive FPC algorithm improves cache efficiency and reduces resource consumption.
IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS
(2022)
Article
Automation & Control Systems
Igor M. L. Pataro, Rita Cunha, Juan D. Gil, Jose L. Guzman, Manuel Berenguel, Joao M. Lemos
Summary: This study introduces an adaptive optimal model-free controller for solar collector fields (SCFs) that overcomes the challenges of using high-complex models. The proposed controller is based on the Reinforcement Q-Learning algorithm and achieves optimal performance using only plant measurements. It outperforms model-based controllers by handling nonlinearities, time-varying model parameters, and computational costs associated with nonlinear models. Simulations using actual data from a thermal plant demonstrate the effectiveness of the model-free controller, as the Q-Learning algorithm converges to the optimal gains of the Linear Quadratic Tracking (LQT) controller.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2023)
Article
Automation & Control Systems
Farzan Soleymani, Md Suruz Miah, Davide Spinello
Summary: Formulated in the framework of Bellman's optimality, area coverage control with multi-agent systems is achieved using an adaptive control policy based on a neural network-based reinforcement learning technique. The method, which relies on local data gathered and shared by each agent, interpolates to achieve optimal configurations consistent with Lloyd's algorithm.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2023)
Article
Computer Science, Artificial Intelligence
Bingjie Ding, Yingnan Pan, Qing Lu
Summary: In this paper, a neural adaptive optimal control strategy is proposed for strict-feedback nonlinear multiagent systems (MASs) with full-state constraints and immeasurable states. The reinforcement learning (RL) with the actor-critic architecture is employed to solve the Hamilton-Jacobi-Bellman (HJB) equation. The introduction of the command filter technique into the value function relaxes the bounded condition of the virtual controller derivative, and the tracking control problem of MASs considering full-state constraints and immeasurable states can be solved without violating constraints.
Article
Computer Science, Artificial Intelligence
Adnan Fayyaz Ud Din, Imran Mir, Faiza Gul, Suleman Mir, Syed Sahal Nazli Alhady, Mohammad Rustom Al Nasar, Hamzah Ali Alkhazaleh, Laith Abualigah
Summary: This research presents a novel control architecture for UAVs, utilizing unconventional reinforcement learning techniques to adapt to the unique design and dynamically respond to changing environments. Nonlinear simulations demonstrate the effectiveness of the proposed approach under different environmental conditions.
Article
Automation & Control Systems
Cong Li, Qingchen Liu, Zhehua Zhou, Martin Buss, Fangzhou Liu
Summary: This article proposes an off-policy risk-sensitive reinforcement learning-based control framework to jointly optimize the task performance and constraint satisfaction in a disturbed environment. The risk-aware value function, constructed using the pseudo control and risk-sensitive input and state penalty terms, is introduced to convert the original constrained robust stabilization problem into an equivalent unconstrained optimal control problem. Simulation results reveal the validity of the proposed control framework.
IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS
(2023)
Article
Automation & Control Systems
Shan Xue, Biao Luo, Derong Liu, Ying Gao
Summary: The article introduces an event-triggered constrained optimal tracking control algorithm using integral reinforcement learning (IRL) which transforms the problem into an optimal regulation one and solves the Hamilton-Jacobi-Bellman equation. The event-triggering mechanism helps ease the pressure on data transmission, making it suitable for control systems with limited computational and communication resources.
INTERNATIONAL JOURNAL OF ROBUST AND NONLINEAR CONTROL
(2021)
Article
Mathematics, Applied
Jingrui Sun
Summary: This paper examines a linear quadratic optimal control problem with fixed terminal states and integral quadratic constraints. By introducing a Riccati equation and utilizing results from duality theory, the optimal control is derived as a target-dependent feedback of the current state.
APPLIED MATHEMATICS AND OPTIMIZATION
(2021)
Article
Engineering, Electrical & Electronic
Yufan Zhang, Honglin Wen, Qiuwei Wu, Qian Ai
Summary: Prediction intervals (PIs) are effective tool for quantifying uncertainty in distribution systems. Traditional central PIs are not suitable for skewed distributions and their offline training is vulnerable to unforeseen changes. We propose an optimal online estimation approach that adapts to different data distributions by adaptively determining probability proportion pairs for quantiles. The approach uses reinforcement learning to integrate adaptive selection and quantile predictions, improving PIs' quality. Case studies show that the proposed method outperforms traditional methods in adapting to data distribution and is more robust against concept drift.
IEEE TRANSACTIONS ON SMART GRID
(2023)
Article
Engineering, Electrical & Electronic
Ataollah Gogani Khiabani, Ali Heydari
IET POWER ELECTRONICS
(2020)
Article
Engineering, Mechanical
Farshid Asadi, Ali Heydari
PROCEEDINGS OF THE INSTITUTION OF MECHANICAL ENGINEERS PART K-JOURNAL OF MULTI-BODY DYNAMICS
(2020)
Article
Computer Science, Information Systems
Ali Heydari, MasoudReza Aghabozorgi, Mehrzad Biguesh
Article
Telecommunications
Ali Heydari, Masoudreza Aghabozorgi
WIRELESS PERSONAL COMMUNICATIONS
(2020)
Article
Food Science & Technology
Ali Heydari, Seyed Mohammad Ali Razavi, Mohammad Ali Hesarinejad, Asgar Farahnaky
Summary: The study investigates the impact of high hydrostatic pressure (HHP) on different types of starch at varying concentrations, revealing that increasing concentration leads to more noticeable starch granule destruction and different resistance levels among starch types. Additionally, HHP treatment and concentration increase significantly affect the viscosity and gelatinization properties of starches.
Article
Food Science & Technology
Ali Heydari, Seyed Mohammad Ali Razavi, Asgar Farahnaky
Summary: The research showed that using HPWS as a fat replacer can improve the stability and physical properties of emulsions, making them more solid-like. Increasing the concentration of HPWS increases the G' value of the emulsions, while increasing the level of fat reduction leads to the opposite effect.
INNOVATIVE FOOD SCIENCE & EMERGING TECHNOLOGIES
(2021)
Article
Biochemistry & Molecular Biology
Ali Heydari, Seyed Mohammad Ali Razavi
Summary: The study showed that high hydrostatic pressure-treated corn starch and waxy corn starch as fat replacers exhibited good stability and rheological characteristics in low-fat emulsions, with waxy corn starch samples showing better performance in terms of elastic modulus and spreadability.
INTERNATIONAL JOURNAL OF BIOLOGICAL MACROMOLECULES
(2021)
Article
Microbiology
Ali Akbar Heydari, Abas Eslami, Maliheh Dadgarmoghaddam
Summary: The study evaluated the clinical characteristics and comorbidities of patients with S. aureus infection to define predictors of adverse outcomes. Vascular catheters and cardiovascular diseases, including hypertension, were found to be common factors associated with S. aureus bacteremia. Different outcomes were seen based on the source of infection and infective endocarditis in these patients, with 61.9% discharged in good condition and 38.1% mortality rate.
JUNDISHAPUR JOURNAL OF MICROBIOLOGY
(2021)
Article
Infectious Diseases
Ali Akbar Heydari, Ehsan Mostafavi, Masoumeh Heidari, Mina Latifian, Saber Esmaeili
Summary: A 60-year-old male farmer and rancher presented symptoms of weight loss, fever, severe sweating, weakness, and anorexia, leading to a diagnosis of chronic Q fever endocarditis. Despite a negative PCR result for C. burnetii in the blood sample, positive IgG antibodies against C. burnetii were detected. Treatment with doxycycline and hydroxychloroquine was successful.
CASE REPORTS IN INFECTIOUS DISEASES
(2021)
Article
Statistics & Probability
Masoud Tavakoli, Ali Akbar Heydari
Summary: In this paper, an economic statistical design for quality characteristics with Exponential distributions is proposed using variable sampling method. The method transforms the Exponential distribution to a Normal distribution and obtains design parameters using an economic model. The results show that the proposed method is more effective than the fixed ratio sampling method when a moderate shift occurs.
COMMUNICATIONS IN STATISTICS-THEORY AND METHODS
(2023)
Article
Nuclear Science & Technology
Rezvan Torkaman, Mehran Heydari, Javad Najafi Cheshmeh, Ali Heydari, Mehdi Asadollahzadeh
Summary: In this investigation, three different frameworks of liquid-liquid extraction (L.L.E) were examined for determining slip velocity (S.V) and characteristic velocity (C.V) in a non-center rotating column (N.C.R.C) with a wide range of factors. Experimental tests were conducted using three dual systems with different interfacial tensions: toluene-water (high), n-butyl acetate-water (medium), and n-butanol-water (low). Two common relationships were proposed for predicting S.V and C.V, involving phase flow rates, rotor speed, column geometry, and physical properties. The proposed relationships were found to be more accurate than those previously reported.
NUCLEAR ENGINEERING AND TECHNOLOGY
(2022)
Article
Microbiology
Ali Akbar Heydari, Shima Jamialahmadi, Mahdi Kouhi Noghondar
Summary: This study evaluated the characteristics of patients with Pseudomonas aeruginosa bloodstream infection (BSI) and compared differences between nosocomial and community-acquired infection (CAI) patients. The results highlighted potential risk factors and clinical symptoms associated with mortality.
JUNDISHAPUR JOURNAL OF MICROBIOLOGY
(2022)
Article
Computer Science, Interdisciplinary Applications
Ali Heydari, Zahra Sayyah Alborzi, Younes Amini, Amin Hassanvand
Summary: The main contribution of this paper is to formulate the problem of optimal design for a renewable wind/solar/biomass hybrid system in Iran and compare the performance of genetic algorithm (GA) and particle swarm optimization (PSO) on this problem. The research on solar and wind hybrid energy systems is abundant, but research on solar/wind/biomass hybrid energy systems is rare. The study finds that the biomass/PV system is the most cost-effective for supplying the required load, and PSO outperforms GA in terms of results.
INTERNATIONAL JOURNAL OF MODERN PHYSICS C
(2023)
Article
Statistics & Probability
Sahera Hussein Zain Al-Thalabi, Ali Akbar Heydari, Masoud Tavakoli
Summary: This study predicts the foreign direct investment of Qatar using artificial neural networks. The multi-layer neural network model gives accurate results and has a strong predictive ability. The feed-forward artificial neural network model is superior to other models and can be used to forecast the GDP of Qatar and predict future investor attraction and market recovery.
JOURNAL OF STATISTICS AND MANAGEMENT SYSTEMS
(2022)
Article
Statistics & Probability
Ali Akbar Heydari, Sahera Hussein Zain Al-Thalabi
Summary: The study applied the binary logistic regression model to analyze binary data and explore the most important factors affecting cancerous tumors. The results showed that, apart from gender, smoking, chronic diseases, family history, weight, height, and marital status were the six significant factors influencing cancer diseases.
JOURNAL OF STATISTICS AND MANAGEMENT SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Hamdan Abdellatef, Lina J. Karam
Summary: This paper proposes performing the learning and inference processes in the compressed domain to reduce computational complexity and improve speed of neural networks. Experimental results show that modified ResNet-50 in the compressed domain is 70% faster than traditional spatial-based ResNet-50 while maintaining similar accuracy. Additionally, a preprocessing step with partial encoding is suggested to improve resilience to distortions caused by low-quality encoded images. Training a network with highly compressed data can achieve good classification accuracy with significantly reduced storage requirements.
Article
Computer Science, Artificial Intelligence
Victor R. Barradas, Yasuharu Koike, Nicolas Schweighofer
Summary: Inverse models are essential for human motor learning as they map desired actions to motor commands. The shape of the error surface and the distribution of targets in a task play a crucial role in determining the speed of learning.
Article
Computer Science, Artificial Intelligence
Ting Zhou, Hanshu Yan, Jingfeng Zhang, Lei Liu, Bo Han
Summary: We propose a defense strategy that reduces the success rate of data poisoning attacks in downstream tasks by pre-training a robust foundation model.
Article
Computer Science, Artificial Intelligence
Hao Sun, Li Shen, Qihuang Zhong, Liang Ding, Shixiang Chen, Jingwei Sun, Jing Li, Guangzhong Sun, Dacheng Tao
Summary: In this paper, the convergence rate of AdaSAM in the stochastic non-convex setting is analyzed. Theoretical proof shows that AdaSAM has a linear speedup property and decouples the stochastic gradient steps with the adaptive learning rate and perturbed gradient. Experimental results demonstrate that AdaSAM outperforms other optimizers in terms of performance.
Article
Computer Science, Artificial Intelligence
Juntong Yun, Du Jiang, Li Huang, Bo Tao, Shangchun Liao, Ying Liu, Xin Liu, Gongfa Li, Disi Chen, Baojia Chen
Summary: In this study, a dual manipulator grasping detection model based on the Markov decision process is proposed. By parameterizing the grasping detection model of dual manipulators using a cross entropy convolutional neural network and a full convolutional neural network, stable grasping of complex multiple objects is achieved. Robot grasping experiments were conducted to verify the feasibility and superiority of this method.
Article
Computer Science, Artificial Intelligence
Miaohui Zhang, Kaifang Li, Jianxin Ma, Xile Wang
Summary: This paper proposes an unsupervised person re-identification (Re-ID) method that uses two asymmetric networks to generate pseudo-labels for each other by clustering and updates and optimizes the pseudo-labels through alternate training. It also designs similarity compensation and similarity suppression based on the camera ID of pedestrian images to optimize the similarity measure. Extensive experiments show that the proposed method achieves superior performance compared to state-of-the-art unsupervised person re-identification methods.
Article
Computer Science, Artificial Intelligence
Florian Bacho, Dominique Chu
Summary: This paper proposes a new approach called the Forward Direct Feedback Alignment algorithm for supervised learning in deep neural networks. By combining activity-perturbed forward gradients, direct feedback alignment, and momentum, this method achieves better performance and convergence speed compared to other local alternatives to backpropagation.
Article
Computer Science, Artificial Intelligence
Xiaojian Ding, Yi Li, Shilin Chen
Summary: This research paper addresses the limitations of recursive feature elimination (RFE) and its variants in high-dimensional feature selection tasks. The proposed algorithms, which introduce a novel feature ranking criterion and an optimal feature subset evaluation algorithm, outperform current state-of-the-art methods.
Article
Computer Science, Artificial Intelligence
Naoko Koide-Majima, Shinji Nishimoto, Kei Majima
Summary: Visual images observed by humans can be reconstructed from brain activity, and the visualization of arbitrary natural images from mental imagery has been achieved through an improved method. This study provides a unique tool for directly investigating the subjective contents of the brain.
Article
Computer Science, Artificial Intelligence
Huanjie Tao, Qianyue Duan
Summary: In this paper, a hierarchical attention network with progressive feature fusion is proposed for facial expression recognition (FER), addressing the challenges posed by pose variation, occlusions, and illumination variation. The model achieves enhanced performance by aggregating diverse features and progressively enhancing discriminative features.
Article
Computer Science, Artificial Intelligence
Zhenyi Wang, Pengfei Yang, Linwei Hu, Bowen Zhang, Chengmin Lin, Wenkai Lv, Quan Wang
Summary: In the face of the complex landscape of deep learning, we propose a novel subgraph-level performance prediction method called SLAPP, which combines graph and operator features through an innovative graph neural network called EAGAT, providing accurate performance predictions. In addition, we introduce a mixed loss design with dynamic weight adjustment to improve predictive accuracy.
Article
Computer Science, Artificial Intelligence
Yiyang Yin, Shuangling Luo, Jun Zhou, Liang Kang, Calvin Yu-Chian Chen
Summary: Medical image segmentation is crucial for modern healthcare systems, especially in reducing surgical risks and planning treatments. Transanal total mesorectal excision (TaTME) has become an important method for treating colon and rectum cancers. Real-time instance segmentation during TaTME surgeries can assist surgeons in minimizing risks. However, the dynamic variations in TaTME images pose challenges for accurate instance segmentation.
Article
Computer Science, Artificial Intelligence
Teng Cheng, Lei Sun, Junning Zhang, Jinling Wang, Zhanyang Wei
Summary: This study proposes a scheme that combines the start-stop point signal features for wideband multi-signal detection, called Fast Spectrum-Size Self-Training network (FSSNet). By utilizing start-stop points to build the signal model, this method successfully solves the difficulty of existing deep learning methods in detecting discontinuous signals and achieves satisfactory detection speed.
Article
Computer Science, Artificial Intelligence
Wenming Wu, Xiaoke Ma, Quan Wang, Maoguo Gong, Quanxue Gao
Summary: The layer-specific modules in multi-layer networks are critical for understanding the structure and function of the system. However, existing methods fail to accurately characterize and balance the connectivity and specificity of these modules. To address this issue, a joint learning graph clustering algorithm (DRDF) is proposed, which learns the deep representation and discriminative features of the multi-layer network, and balances the connectivity and specificity of the layer-specific modules through joint learning.
Article
Computer Science, Artificial Intelligence
Guanghui Yue, Guibin Zhuo, Weiqing Yan, Tianwei Zhou, Chang Tang, Peng Yang, Tianfu Wang
Summary: This paper proposes a novel boundary uncertainty aware network (BUNet) for precise and robust colorectal polyp segmentation. BUNet utilizes a pyramid vision transformer encoder to learn multi-scale features and incorporates a boundary exploration module (BEM) and a boundary uncertainty aware module (BUM) to handle boundary areas. Experimental results demonstrate that BUNet outperforms other methods in terms of performance and generalization ability.