Recent Advances in Deep Reinforcement Learning Applications for Solving Partially Observable Markov Decision Processes (POMDP) Problems: Part 1—Fundamentals and Applications in Games, Robotics and Natural Language Processing

Abstract

The first part of a two-part series of papers provides a survey on recent advances in Deep Reinforcement Learning (DRL) applications for solving partially observable Markov decision processes (POMDP) problems. Reinforcement Learning (RL) is an approach to simulate the human’s natural learning process, whose key is to let the agent learn by interacting with the stochastic environment. The fact that the agent has limited access to the information of the environment enables AI to be applied efficiently in most fields that require self-learning. Although efficient algorithms are being widely used, it seems essential to have an organized investigation—we can make good comparisons and choose the best structures or algorithms when applying DRL in various applications. In this overview, we introduce Markov Decision Processes (MDP) problems and Reinforcement Learning and applications of DRL for solving POMDP problems in games, robotics, and natural language processing. A follow-up paper will cover applications in transportation, communications and networking, and industries.

Application of Production-Oriented Approach and Deep Learning Model in English Translation Teaching Model

“…Table 6 shows that the average score for volume B of the translating ability exam is 26.70 and the average score for volume B is 25. 14. is shows that after three months of translation education experiment, the average value of experimental translation ability is 1.56 points higher, with little difference.…”

“…e experiment shows that POA is a typical application based on design research, and that it can be extended to other educational contexts and to beginners by expanding its scope, geographical area, and target audience. [14] reflects on the motivating, enabling, and evaluating aspects of POA. In contrast, [15] argues that the design of the three components of POA promotes students' curiosity and weakens teachers' role of "scaffolding," which has a positive effect on students' independent learning.…”

Exploring the Application of the Random Matrix Thinking Model in Teaching English Predicate Constructions

“…Although it is a simple method, it can cause some impact on the performance by incorrectly processing nouns due to the lack of rigor in the processing of noun predicates. Statistically based, various features are set in the Co Nll2008 corpus, such as Xiang and Foo using the same qualities in predicate recognition and word sense credit and getting better results [20]. For example, Reisman et al proposed more and more valuable features to expand further the feature set of predicate recognition, which led to a significant improvement in the performance of predicate recognition [21].…”

Comparative Analysis of Routing Schemes Based on Machine Learning

“…e Google DeepMind institute proposed deep Q-learning (DQN) [43]. It uses a deep neural network (DNN) instead of the original Q-value table to approximate the Q-function and trains it through the squared error:…”