Deep transfer reinforcement learning for resource allocation in hybrid multiple access systems

Through well-designed optimization paradigm and deep neural networks as feature extractor, deep reinforcement learning (DRL) algorithms learn optimal policy on discrete and continuous action space. However, such capability is restricted by the low sampling efficiency. By inspecting the importance of feature extraction in DRL, we find that state feature learning is one of the key obstacles for sampling efficiently. To this end, we propose a new state representation learning scheme with adjacent state consistency loss (ASC loss). The loss is based on the hypothesis that the distance between adjacent states is smaller than that of far apart ones since scenes in videos generally evolve smoothly. We exploit ASC loss as an assistant of RL loss in the training phase to boost the state feature learning, and make evaluation on existing DRL algorithms as well as behavioral cloning algorithm. Experiments on Atari games and MuJoCo continuous control tasks demonstrate the effectiveness of our scheme.

State Representation Learning With Adjacent State Consistency Loss for Deep Reinforcement Learning

Swimming monitoring based on acceleration sensor is an emerging research direction in the field of human motion recognition. As a public sport, swimming has a wide audience. The swimming monitoring system can facilitate people to monitor and record their own swimming data, so as to formulate a reasonable training plan. Aiming at the defects of single modal information representation ability, high contingency, and easy to be influenced by the outside world, this paper adopts the underwater posture training model of swimmers to perform multimodal information fusion. In this paper, a multimodal information fusion method based on evolutionary neural network is proposed, and an intelligent perception information processing model of the intelligent subject system is constructed. Aiming at the defect that the accuracy and speed of the underwater posture monitoring of swimmers cannot be guaranteed in a complex environment, an evolutionary neural network optimized by a multimodal adaptive genetic algorithm is constructed to perform multimodal information fusion to ensure the effectiveness of the system in the face of complex information. Regarding attitude detection, it mainly uses the three dimensions of the angle of movement, the influence of gravity, and the strength and speed of the movement to measure. The MPU6050 module processor has a wide range of applications and is a mold processing tool with high performance and level. It completes the data processing, data calculation, and data storage of the inspection system in this paper. This paper further studies the working principle, structure, and operation process of this module and adjusts the time error in the detection of carrier motion and attitude so that the processing function of this module can play an optimal state. Four kinds of swimming posture measurement experiments were carried out on the swimmers, and the experimental data were analyzed. The whole system is controlled by the host computer man-machine interaction software remotely and in real time through commands. The experimental results show that the system realizes the detection of the basic posture, meets the basic requirements of the system design, and provides a certain foundation for the follow-up research.

/pdf/construction-of-swimmer-s-underwater-posture-training-model-24b01r27.pdf

Construction of Swimmer's Underwater Posture Training Model Based on Multimodal Neural Network Model

Rule of thirds-aware reinforcement learning for image aesthetic cropping

Reinforcement Learning (RL) is a machine learning (ML) technique to learn sequential decision-making in complex problems. RL is inspired by trial-and-error based human/animal learning. It can learn an optimal policy autonomously with knowledge obtained by continuous interaction with a stochastic dynamical environment. Problems considered virtually impossible to solve, such as learning to play video games just from pixel information, are now successfully solved using deep reinforcement learning. Without human intervention, RL agents can surpass human performance in challenging tasks. This review gives a broad overview of RL, covering its fundamental principles, essential methods, and illustrative applications. The authors aim to develop an initial reference point for researchers commencing their research work in RL. In this review, the authors cover some fundamental model-free RL algorithms and pathbreaking function approximation-based deep RL (DRL) algorithms for complex uncertain tasks with continuous action and state spaces, making RL useful in various interdisciplinary fields. This article also provides a brief review of model-based and multi-agent RL approaches. Finally, some promising research directions for RL are briefly presented. 

Reinforcement Learning Algorithms: A brief survey

Recent years have witnessed the great success of deep reinforcement learning (DRL) on a variety of vision games. Although DNN has demonstrated strong power in representation learning, such capacity is under-explored in most DRL works whose focus is usually on optimization solvers. In fact, we discover that the state feature learning is the main obstacle for further improvement of DRL algorithms. To address this issue, we propose a new state representation learning scheme with our Adjacent State Consistency Loss (ASC Loss). The loss is defined based on the hypothesis that there are fewer changes between adjacent states than that of far apart ones, since scenes in videos generally evolve smoothly. In this paper, we exploit ASC loss as an assistant of RL loss in the training phase to boost the state feature learning. We conduct evaluation on Atari games and MuJoCo continuous control tasks, which demonstrates that our method is superior to OpenAI baselines.

Tianyu Zhao

Papers

State Representation Learning With Adjacent State Consistency Loss for Deep Reinforcement Learning

State Representation Learning For Effective Deep Reinforcement Learning