America is becoming more divided between its rural and urban areas on a political and economic level. By boosting resilience to economic shocks, generating jobs, and quickening economic growth, entrepreneurship can assist rural communities in catching up. In contrast to metropolitan areas, there is little information available about business formation in rural areas. This research proposes novel technique in cooperation strategies based on rural small business entrepreneurship using game theory with machine learning model in conflict optimization. Here the small business entrepreneurship strategy analysis using support vector adversarial gradient game theory model. then the optimization has been carried using multi agent regression convolutional Q-learning. Development method of the current rural industries is recorded from standpoint of entrepreneurship, global classic practical experience of digitising rural industries is investigated. F-1 score, average precision, recall, efficiency, and prediction accuracy have all been examined through experimental study. Using machine learning, we predict small businesses' active participation in community resilience. We find that performance of model is significantly enhanced by addition of social capital predictors. The proposed model attained prediction accuracy of 97%, F-1 score of 95%, average precision of 93%, recall of 92%, efficiency of 94%.