Abstract:

This study explores the application of Reinforcement Learning models, specifically Deep Q-Networks (DQN), for optimizing distribution strategies in the gaming industry. The analyzed data includes key variables such as the estimated number of owners, the price of the game, the number of reviews and the recommendations. The data was pre-processed and normalized to ensure the stability of the model.

The developed DQN model was trained using adaptive strategies and evaluated using key metrics such as Root Mean Squared Error (RMSE) and Mean Absolute Percentage Error (MAPE). These metrics demonstrate the high predictive accuracy of the model and underline its ability to recognize complex patterns in the data and optimize distribution strategies. The visualization of the results, such as the distribution of recommendations and the relationship between price and recommendations after data transformation, further underscores the effectiveness of the approach.

This study emphasizes the potential of the Deep Q-Network model in optimizing distribution strategies, with a particular focus on real-time adaptation, enhancing user engagement, and improving market performance.