This paper provides a comprehensive overview of the applications of game theory in deep learning. Today, deep learning is a fast-evolving area for research in the domain of artificial intelligence. Alternatively, game theory has been showing its multi-dimensional applications in the last few decades. The application of game theory to deep learning includes another dimension in research. Game theory helps to model or solve various deep learning-based problems. Existing research contributions demonstrate that game theory is a potential approach to improve results in deep learning models. The design of deep learning models often involves a game-theoretic approach. Most of the classification problems which popularly employ a deep learning approach can be seen as a Stackelberg game. Generative Adversarial Network (GAN) is a deep learning architecture that has gained popularity in solving complex computer vision problems. GANs have their roots in game theory. The training of the generators and discriminators in GANs is essentially a two-player zero-sum game that allows the model to learn complex functions. This paper will give researchers an extensive account of significant contributions which have taken place in deep learning using game-theoretic concepts thus, giving a clear insight, challenges, and future directions. The current study also details various real-time applications of existing literature, valuable datasets in the field, and the popularity of this research area in recent years of publications and citations.

Purpose The purpose of this paper is to propose a model for a target searching problem in a two-dimensional region with time constraints. The proposed model facilitates the search operation by minimizing the mission time and fuel usage, and the search operation is performed by a set of agents divided into a number of groups. Design/methodology/approach The authors have applied optimization techniques, Cartesian product, inclusion–exclusion principle, cooperative strategy, Shapley value, fuzzy Shapley function and Choquet integral to model the problem. Findings The proposed technique optimizes the placement of base stations that minimizes the sortie length of the agents. The results show that the cooperative strategy outperforms the non-cooperative strategy. The Shapley values quantify the rewards of each group based on their contributions to the search operation, whereas the fuzzy Shapley values determine the rewards of each group based on their contributions and level of cooperation in the search operation. Practical implications The proposed model can be applied to model many real-time problems such as patrolling in international borders, urban areas, forests and managing rescue operations after natural calamities, etc. Therefore, defence organizations, police departments and other operation management sectors will be benefitted by applying the proposed approach. Originality/value To the best of the authors’ knowledge, determining the optimal locations of base stations in a region is not explored in the existing works on target searching problems with fuel constraints. The proposed approach to cooperatively search the targets in a region is new. Introducing the Shapley function and fuzzy Shapley function is a novel idea to quantify the rewards of each group based on their contributions and level of cooperation in the search operation. This paper addresses these unexplored areas.

Abstract-Target searching is one of the challenging research areas in defense. Different types of sensor networks are deployed for searching targets in critical zones. The selection of optimal strategies for the sensor nodes under certain constraints is the key issue in target searching problem. This paper addresses a number of target searching problems related to various defense scenarios and introduces new strategic approaches to facilitate the search operation for the mobile sensors in a two-dimensional bounded space. The paper classifies the target searching problems into two categories: preference-based and traversal distance based. In the preference based problems, the strategies for the mobile sensors are determined by Stable Marriage Problem, College Admission Problem, and voting system; they are analyzed with suitable examples. Alternatively, traversal distance based problems are solved by our proposed graph searching approaches and analyzed with randomly chosen examples. Results obtained from the examples signify that our proposed models can be applied in defense-related target searching problems.

No abstract

No abstract