On Maxwell Electrodynamics in Multi-Dimensional Spaces

The governing equations of Maxwell electrodynamics in multi-dimensional spaces are derived from the variational principle of least action, which is applied to the action function of the electromagnetic field. The Hamiltonian approach for the electromagnetic field in multi-dimensional pseudo-Euclidean (flat) spaces has also been developed and investigated. Based on the two arising first-class constraints, we have generalized to multi-dimensional spaces a number of different gauges known for the three-dimensional electromagnetic field. For multi-dimensional spaces of non-zero curvature the governing equations for the multi-dimensional electromagnetic field are written in a manifestly covariant form. Multi-dimensional Einstein’s equations of metric gravity in the presence of an electromagnetic field have been re-written in the true tensor form. Methods of scalar electrodynamics are applied to analyze Maxwell equations in the two and one-dimensional spaces.

Diffusion treatment of quantum mechanics and its consequences

Localized ensemble of free microparticles spreads out as in a frictionless diffusion satisfying the principle of relativity. An ensemble of classical particles in a fluctuating classical scalar field diffuses in a similar way, and this analogy is used to formulate diffusion quantum mechanics (DQM). DQM reproduces quantum mechanics for homogeneous and gravity for inhomogeneous scalar field. Diffusion flux and probability density are related by Fick’s law, diffusion coefficient is constant and invariant. Hamiltonian includes a “thermal” energy, kinetic energies of drift and diffusion flux. The probability density and the action function of drift form a canonical pair and canonical equations for them lead to the Hamilton-Jacobi-Madelung and continuity equations. At canonical transformation to a complex probability amplitude they form a linear Schrödinger equation. DQM explains appearance of quantum statistics, rest energy (“thermal” energy) and gravity (“thermal” diffusion) and leads to a low mass mechanism for composite particles.

Boosting Offline Reinforcement Learning with Residual Generative Modeling

Offline reinforcement learning (RL) tries to learn the near-optimal policy with recorded offline experience without online exploration.Current offline RL research includes: 1) generative modeling, i.e., approximating a policy using fixed data; and 2) learning the state-action value function. While most research focuses on the state-action function part through reducing the bootstrapping error in value function approximation induced by the distribution shift of training data, the effects of error propagation in generative modeling have been neglected. In this paper, we analyze the error in generative modeling. We propose AQL (action-conditioned Q-learning), a residual generative model to reduce policy approximation error for offline RL. We show that our method can learn more accurate policy approximations in different benchmark datasets. In addition, we show that the proposed offline RL method can learn more competitive AI agents in complex control tasks under the multiplayer online battle arena (MOBA) game, Honor of Kings.

Energy Analysis of the Propulsion Shaft Fatigue Process in a Rotating Mechanical System Part III Dimensional Analysis

This article presents the third and last part of the problem of diagnosing the fatigue of marine propulsion shafts in terms of energy with the use of the action function, undertaken by the authors. Even the most perfect physical models of real objects, observed under laboratory conditions and developed based on the results of their research, cannot be useful in diagnostics without properly transferring the obtained results to the scale of the real object. This paper presents the method of using dimensional analyses and the Buckingham theorem (the so-called π theorem) to determine the dimensionless numbers of the dynamic similarity of the physical model of the propulsion shaft and its real ship counterpart, which enable the transfer of the results of the research on the energy processes accompanying the ship propulsion shaft fatigue from the physical model to the real object.

Comparing Action Aggregation Strategies in Deep Reinforcement Learning with Continuous Action

Deep Reinforcement Learning has been very promising in learning continuous control policies. For complex tasks, Reinforcement Learning with minimal human intervention is still a challenge. This article proposes a study to improve performance and to stabilize the learning curve using the ensemble learning methods. Learning a combined parameterized action function using multiple agents in a single environment, while searching for a better way to learn, regardless of the quality of the parametrization. The action ensemble methods were applied in three environments: pendulum swing-up, cart pole and half cheetah. Their results demonstrated that action ensemble can improve performance with respect to the grid search technique. This article also presents as contribution the comparison of the effectiveness of the aggregation techniques, the analysis considers the use of the separate or the combined policies during training. The latter presents better learning results when used with the data center aggregation strategy.

Farmer cooperatives, gender and side-selling behavior in Burundi's coffee sector

PurposeThis study aims to improve our understanding of side selling in farmer cooperatives. Cooperative member side selling, in which farmers divert produce from cooperatives to competitors, threatens coffee cooperatives. This is a problem in Burundi, where many households earn income from coffee and cooperatives serve a collective action function.Design/methodology/approachUsing data from a survey of Burundian coffee farmers, we assess the determinants of two types of cooperative member side-selling behavior: selling to both cooperative and non-cooperative buyers and selling solely to non-cooperative buyers.FindingsFarmers who sell to both cooperative and non-cooperative buyers are more likely to be male household heads, be more invested in coffee and have larger farms than non-side sellers, among other characteristics. Farmers who only sell to non-cooperative buyers are poorer and less invested in coffee than non-side sellers.Research limitations/implicationsAdditional research is needed to better understand why side-selling behavior differs between groups and to better understand how household head gender influences side selling. In addition, this study lacks qualitative data supporting quantitative findings. Future research should include qualitative methods to better understand motivations for side-selling behavior.Originality/valueThe study provides important information on what influences cooperative member side selling and focuses on specific types of side-selling behavior that have been largely overlooked. The study focuses on the role of household head gender in side selling, which is important, given the centrality of women to African agriculture.

MORPHOLOGY OF FOLKLORE BATU NAGA LAMPUNG

The purpose of this study was to analyze the morphology of folklore based on Vladimir Propp's theory and method of researching Narratology structure including the structure of the actor's action function, the function in the action environment, and the scheme. This research method is descriptive qualitative and the method developed by Vladimir Propp. The object of research is the Batu Naga Lampung folklore. The technique of data collection is done by the documentation technique through a literature study. Data analysis techniques use the flow method that is data reduction, data presentation, and concluding. The results of the study found there are 26 functions of the perpetrators' actions that are distributed into 6 action environments.

Multi-agent cooperation Q-learning algorithm based on constrained Markov Game

Multi-Agent system has broad application in real world, whose security performance, however, is barely considered. Reinforcement learning is one of the most important methods to resolve Multi-Agent problems. At present, certain progress has been made in applying Multi-Agent reinforcement learning to robot system, man-machine match, and automatic, etc. However, in the above area, an agent may fall into unsafe states where the agent may find it difficult to bypass obstacles, to receive information from other agents and so on. Ensuring the safety of Multi-Agent system is of great importance in the above areas where an agent may fall into dangerous states that are irreversible, causing great damage. To solve the safety problem, in this paper we introduce a Multi-Agent Cooperation Q-Learning Algorithm based on Constrained Markov Game. In this method, safety constraints are added to the set of actions, and each agent, when interacting with the environment to search for optimal values, should be restricted by the safety rules, so as to obtain an optimal policy that satisfies the security requirements. Since traditional Multi-Agent reinforcement learning algorithm is no more suitable for the proposed model in this paper, a new solution is introduced for calculating the global optimum state-action function that satisfies the safety constraints. We take advantage of the Lagrange multiplier method to determine the optimal action that can be performed in the current state based on the premise of linearizing constraint functions, under conditions that the state-action function and the constraint function are both differentiable, which not only improves the efficiency and accuracy of the algorithm, but also guarantees to obtain the global optimal solution. The experiments verify the effectiveness of the algorithm.

Damaged buildings recognition of post-earthquake high-resolution remote sensing images based on feature space and decision tree optimization

Multi-Agent system has broad application in real world, whose security performance, however, is barely considered. Reinforcement learning is one of the most important methods to resolve Multi-Agent problems. At present, certain progress has been made in applying Multi-Agent reinforcement learning to robot system, man-machine match, and automatic, etc. However, in the above area, an agent may fall into unsafe states where the agent may find it difficult to bypass obstacles, to receive information from other agents and so on. Ensuring the safety of Multi-Agent system is of great importance in the above areas where an agent may fall into dangerous states that are irreversible, causing great damage. To solve the safety problem, in this paper we introduce a Multi-Agent Cooperation Q-Learning Algorithm based on Constrained Markov Game. In this method, safety constraints are added to the set of actions, and each agent, when interacting with the environment to search for optimal values, should be restricted by the safety rules, so as to obtain an optimal policy that satisfies the security requirements. Since traditional Multi-Agent reinforcement learning algorithm is no more suitable for the proposed model in this paper, a new solution is introduced for calculating the global optimum state-action function that satisfies the safety constraints. We take advantage of the Lagrange multiplier method to determine the optimal action that can be performed in the current state based on the premise of linearizing constraint functions, under conditions that the state-action function and the constraint function are both differentiable, which not only improves the efficiency and accuracy of the algorithm, but also guarantees to obtain the global optimal solution. The experiments verify the effectiveness of the algorithm.

Fuzzy Reinforcement Learning and Curriculum Transfer Learning for Micromanagement in Multi-Robot Confrontation

Multi-Robot Confrontation on physics-based simulators is a complex and time-consuming task, but simulators are required to evaluate the performance of the advanced algorithms. Recently, a few advanced algorithms have been able to produce considerably complex levels in the context of the robot confrontation system when the agents are facing multiple opponents. Meanwhile, the current confrontation decision-making system suffers from difficulties in optimization and generalization. In this paper, a fuzzy reinforcement learning (RL) and the curriculum transfer learning are applied to the micromanagement for robot confrontation system. Firstly, an improved Qlearning in the semi-Markov decision-making process is designed to train the agent and an efficient RL model is defined to avoid the curse of dimensionality. Secondly, a multi-agent RL algorithm with parameter sharing is proposed to train the agents. We use a neural network with adaptive momentum acceleration as a function approximator to estimate the state-action function. Then, a method of fuzzy logic is used to regulate the learning rate of RL. Thirdly, a curriculum transfer learning method is used to extend the RL model to more difficult scenarios, which ensures the generalization of the decision-making system. The experimental results show that the proposed method is effective.