A5100330628- Evolutionary Algorithms and Applications
- Advanced Manufacturing and Logistics Optimization
- Metaheuristic Optimization Algorithms Research
- Industrial Vision Systems and Defect Detection
- Reinforcement Learning in Robotics
- Scheduling and Optimization Algorithms
- Robotic Path Planning Algorithms
- Digital Transformation in Industry
- Manufacturing Process and Optimization
- Target Tracking and Data Fusion in Sensor Networks
- Fuzzy Logic and Control Systems
- Image Processing and 3D Reconstruction
- Advanced Vision and Imaging
- Image and Object Detection Techniques
- Internet Traffic Analysis and Secure E-voting
- Advanced Multi-Objective Optimization Algorithms
- Image Processing Techniques and Applications
- Viral Infectious Diseases and Gene Expression in Insects
- Surface Roughness and Optical Measurements
- Artificial Intelligence in Games
- Machine Learning and Data Classification
- Network Security and Intrusion Detection
- Gaussian Processes and Bayesian Inference
- Adversarial Robustness in Machine Learning
- Radiomics and Machine Learning in Medical Imaging
Fudan University
2023-2025
University of Canterbury
2021-2024
Xiangtan University
2024
Cardiff University
2024
Zhejiang University of Science and Technology
2023
Jiangsu University
2023
Jiangsu Academy of Agricultural Sciences
2023
Zhejiang University of Technology
2022
BOE Technology Group (China)
2021
Seagate (United States)
2021
This paper proposes a self-learning Monte Carlo tree search algorithm (SL-MCTS), which has the ability to continuously improve its problem-solving in single-player scenarios. SL-MCTS combines MCTS with two-branch neural network (PV-Network). The architecture can balance for exploration and exploitation. PV-Network replaces rollout process of predicts promising direction value nodes, increases convergence speed efficiency. an effective method assess trajectory current model during by...
Learning Classifier Systems (LCSs) are a group of rule-based evolutionary computation techniques, which have been frequently applied to data-mining tasks. Evidence shows that LCSs can produce models containing human-discernible patterns. But, traditional cannot efficiently discover consistent, general rules - especially in domains unbalanced class distribution. The reason is search methods, e.g. crossover, mutation, and roulette wheel deletion, rely on stochasticity find keep optimum rules....
The robot path planning for forming complex parts is one of the key techniques in coaxial laser cladding powder. In this paper, a method automatic generation from CAD (computer aided design) proposed. size information and topological model are extracted. redundant vertexes removed. new ordered traversed by binary tree algorithm model. relations between vertexes, edges faces established. displayed with OpenGL. comparison analysis contour hyperbolic curve completed. simulation shows...
War chess gaming has so far received insufficient attention but is a significant component of turn-based strategy games (TBS) and studied in this paper. First, common game model proposed through various existing war types. Based on the model, we propose theory frame involving combinational optimization one hand tree search other hand. We also discuss key problem, namely, that number branching factors each turn huge. Then, two algorithms for searching to solve problem: (<mml:math...
In this paper, we propose a new ant colony optimization algorithm, called learning-based neural (LN-ACO), which incorporates an "intelligent ant". This intelligent contains convolutional network pre-trained on large set of instances is able to predict the selection probabilities possible choices at each step algorithm. The capable generating solution based knowledge learned during training, but also guides other 'traditional' ants in improving their search. As search progresses, influenced...
To assist the decision makers, we develop a new supply chain simulation software: Easy-SC, Java-based tool that simplifies simulation. In its current state of development, Easy-SC is modeling for assessing pros and cons facility locations, resource allocations different combinations policies. It can be used in small projects such as single inventory units to large-scale world wide chains. This paper introduces by an examination software module architecture, elements, basic features processes.
Path redirection for virtual reality (VR) navigation allows the user to explore a large environment (VE) while VR application is hosted in limited physical space. Static mapping methods deform scene fit The challenge reasonable way, making distortions friendly user's visual perception. In this paper we propose feature-guided path method that finds and takes into account features of 3D scenes. first offline step, collection view-independent view-dependent VE are extracted stored feature map....
This paper considers the problem of fixed-interval smoothing for Markovian switching systems with multiple linear state-space models. An enhanced algorithm that is capable accurately approximating Bayesian optimal smoother proposed. It utilizes exact expression quotient two Gaussian densities to help solve backward-time recursive equations smoothing, and computes joint posterior state vector model index. The proposed only involves approximation each model-matched posterior, which a mixture,...
Recently, there are many problems caused by global environment warming. The limited natural resources require efficient methods and systems for recycling processing of the wastes a better environment. One today is kitchen garbage, because when they become in large amounts small areas, capacity cannot assimilate them. Therefore, construction needed to save minimize wastes. For this reason, paper, authors have proposed implemented waste management robot system, which can change garbage...
In this paper, we propose an approach for Chinese accent identification using both cepstral and prosodic features with gender-dependent model. We exploit a combination of conventional Shifted Delta Cepstrum (SDC) pitch contour as example segmental suprasegmental features, to capture the characteristics in accents. use cubic polynomials estimate segments order model differences within train GMM acoustic models express deal gender variation. Since criterion assumption cannot solve those...
Abstract In the multi‐variety and large‐scale order production mode, enterprises must balance delivery deadlines maintain customer satisfaction while also considering health status of machines. Therefore, authors propose a method for jointly optimising scheduling machine maintenance. Before processing, an value grading sorting model health‐status group partitioning are constructed to classify orders into different levels machines groups, respectively. During based on Weibull distribution...
The paper investigates the feasibility of using GANs to create realistic induction motor thermal RGB image datasets for multimodal condition-monitoring systems. Generating high-quality images presents computational challenges, and in this study, two GAN frameworks, DCGAN WGAN-GP, were used under different health conditions. Firstly, was on three conditions various hyperparameters, but results required further improvement. Secondly, WGAN-GP with an extensive training duration 11 hours,...