Corporate social responsibility (CSR) represents a form of international private business self-regulation that contributes to the social aims of a charitable, activist, or philanthropic nature through helping ethically based practices. The benefits from CSR like generating better public image, enhance in community support, enhance in market share, enhance in the satisfaction of the consumer, enhance in the exports, offering better quality services and products, and enhance in the productivity is not considered as the perception level consumers. This paper analyzes the effect of CSR activities on the fast moving consumer goods (FMCG) industries on the basis of the retention of the consumers of FMCG products. Initially, a questionnaire is prepared that consists of five categories. Each category is composed of more than ten questions. The information related to the questionnaire is obtained in the view of the professional students from Odisha. The customers are requested to fill the exact information as much as possible. Next, a higher-order statistical analysis is considered for realizing the behaviour of the categories. Additionally, the usage of deep learning called Optimized RNN (O-RNN) is used for predicting the customer retention in maintaining the CSR on FMCG industries. The improvement in the traditional RNN is done by optimizing the hidden neurons by the improved form of Sea Lion Optimization Algorithm (slno) called Modified slno (M-slno) with the intention of minimizing the error among the actual and the predicted outcome, hence it is called as O-RNN. The retention of the customers regarding the effect of CSR activities on FMCG companies and the benefits from CSR is investigated using the O-RNN prediction analysis against several traditional deep learning models as well as different statistical measures.

企业社会责任 (CSR) 代表了一种国际私营企业自律形式，通过帮助符合道德的实践，有助于实现慈善、激进或慈善性质的社会目标。企业社会责任的好处，如塑造更好的公众形象、提高社区支持、提高市场份额、提高消费者满意度、提高出口、提供更优质的服务和产品以及提高生产力，不被视为认知层面的消费者。本文以快速消费品消费者的留存为基础，分析企业社会责任活动对快速消费品行业的影响。最初，准备一份由五个类别组成的调查问卷。每个类别由十多个问题组成。问卷相关信息是从奥里萨邦专业学生的角度获得的。请客户尽可能填写准确的信息。接下来，考虑更高阶的统计分析来实现类别的行为。此外，使用称为优化 RNN (O-RNN) 的深度学习来预测在维持快速消费品行业的 CSR 中的客户保留率。传统 RNN 的改进是通过改进形式的海狮优化算法 (slno) 优化隐藏神经元来完成的，称为修改 slno (M-slno)，目的是最小化实际结果和预测结果之间的误差，因此它称为 O-RNN。