Zhenbo Cheng

Prof. Dr. Zhenbo Cheng

orcid

Research Focus & Metrics

Academic Career Length

22Years (2005-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
4 January 2018 in PLOS Computational Biology
Verified
Journal Article
Open access
PDF file
ZZZhewei Zhang
Zhewei Zhang
ZCZhenbo Cheng
Zhenbo Cheng
ZLZhongqiao Lin
Zhongqiao Lin
CNChechang Nie
Chechang Nie
TYTianming Yang
Tianming Yang
Reinforcement learning has been widely used in explaining animal behavior. In reinforcement learning, the agent learns the value of the states in the task, collectively constituting the task state space, and uses the knowledge to choose actions and acquire desired outcomes. It has been proposed that the orbitofrontal cortex (OFC) encodes the task state space during reinforcement learning. However, it is not well understood how the OFC acquires and stores task state information. Here, we propose a neural network model based on reservoir computing. Reservoir networks exhibit heterogeneous and dynamic activity patterns that are suitable to encode task states. The information can be extracted by a linear readout trained with reinforcement learning. We demonstrate how the network acquires and stores task structures. The network exhibits reinforcement learning behavior and its aspects resemble experimental findings of the OFC. Our study provides a theoretical explanation of how the OFC may contribute to reinforcement learning and a new approach to understanding the neural mechanism underlying reinforcement learning. Many studies employ reinforcement learning models to explain human or animal behavior, in which it is assumed that the animals know the task structure. Yet in the real life, the task structure also has to be acquired through learning. The orbitofrontal cortex has been proposed to play important roles in representing task structure, yet it is poorly understood how it does it and why it can do it. Here, we use a neural network model based on reservoir computing to approach these questions. We show that it is critical for the network to receive reward information as part of its input. Just as the orbitofrontal cortex that receives converging sensory and reward inputs, the network is able to acquire task structure and support reinforcement learning by encoding a combination of sensory and reward events. The importance of reward inputs in the model explains the sophisticate representations of reward information in the orbitofrontal cortex and provides a theoretic account of the current experimental data.

1 January 2015 in Ecotoxicology and Environmental Safety
Verified
Journal Article
Closed access
GXGang Xiao
Gang Xiao
MFMin Feng
Min Feng
ZCZhenbo Cheng
Zhenbo Cheng
MZMeirong Zhao
Meirong Zhao
JMJiafa Mao
Jiafa Mao
LMLuke Mirowski
Luke Mirowski

1 June 2016
Verified
Conference Paper
Closed access
GXGang Xiao
Gang Xiao
WFWei-Kang Fan
Wei-Kang Fan
JMJiafa Mao
Jiafa Mao
ZCZhenbo Cheng
Zhenbo Cheng
DZDan-Hong Zhong
Dan-Hong Zhong
YLYi Li
Yi Li
Using fish behavior to reflect changes of water quality and aim at water quality warning is currently a popular topic of environmental monitoring. The fish tracking technology with occlusion is the basis for the analysis of fish school behavior, and only tracking each individual fish accurately in the shoal can get the exact quantitative analysis of their feature. In this paper, we design the data acquisition platform by using a plane mirror and a monocular camera, and present a fish tracking with occlusion method by using the three-dimensional information. According to the principle of plane mirror imaging and the camera imaging, we find the relationship between the target in mirror-view and the one in the direct-view. Based on the relationship and underwater imaging process, we calculate the 3D coordinates-for both real fish and imaginary fish under circumstance of occlusion. Finally, we use the 3D coordinate to achieve the target occlusion tracking for fish school. The experimental results show that this approach can obtain precise tracking of three fish and get the 3D coordinate accurately even they are in occlusion.

30 March 2023 in Expert Systems with Applications
Verified
Journal Article
Closed access
JLJiawei Lu
Jiawei Lu
JZJiahong Zheng
Jiahong Zheng
ZCZhenbo Cheng
Zhenbo Cheng
QWQibing Wang
Qibing Wang
DLDuanni Li
Duanni Li
GXGang Xiao
Gang Xiao

11 March 2025 in Mathematics
Verified
Journal Article
Open access
PDF file
JLJiansha Lu
Jiansha Lu
佳张佳蕊 张
佳蕊 张
JCJUN CAO
JUN CAO
XXXuesong Xu
Xuesong Xu
Yiping ShaoYiping Shao
Yiping Shao
ZCZhenbo Cheng
Zhenbo Cheng
In order to address the impact of equipment fault diagnosis and repair delays on production schedule execution in the dynamic scheduling of flexible job shops, this paper proposes a multi-resource, multi-objective dynamic scheduling optimization model, which aims to minimize delay time and completion time. It integrates the scheduling of the workpieces, machines, and maintenance personnel to improve the response efficiency of emergency equipment maintenance. To this end, a self-learning Ant Colony Algorithm based on deep reinforcement learning (ACODDQN) is designed in this paper. The algorithm searches the solution space by using the ACO, prioritizes the solutions by combining the non-dominated sorting strategies, and achieves the adaptive optimization of scheduling decisions by utilizing the organic integration of the pheromone update mechanism and the DDQN framework. Further, the generated solutions are locally adjusted via the feasible solution optimization strategy to ensure that the solutions satisfy all the constraints and ultimately generate a Pareto optimal solution set with high quality. Simulation results based on standard examples and real cases show that the ACODDQN algorithm exhibits significant optimization effects in several tests, which verifies its superiority and practical application potential in dynamic scheduling problems.

28 June 2016 in Lecture Notes in Computer Science
Verified
Book Chapter
Closed access
PDF file
GXGang Xiao
Gang Xiao
TSTengfei Shao
Tengfei Shao
TZTianqi Zhu
Tianqi Zhu
YLYi Li
Yi Li
JMJiafa Mao
Jiafa Mao
ZCZhenbo Cheng
Zhenbo Cheng