Abstract
In contemporary times, the video game industry has experienced remarkable growth, captivating a wide audience with its immersive offerings. Undoubt- edly, it stands as a significant global contributor to revenue generation. This sector wields a considerable influence, drawing in individuals with sharp and innovative skills to foster the expansion of video games worldwide. Explor- ing the substantial profit generated this sector, machine learning technologies have become instrumental in creating highly effective models that can anal- yse and forecast computer game sales well in advance. The realm of machine learning offers a diverse array of models for predicting future sales, employ- ing techniques such as Linear and Multiple Regression, Random Forest, Deci- sion Trees, Support Vector Machines, among others. Each of these approaches processes data using various mathematical concepts and formulas to estimate sales. The selection of an appropriate model depends on a thorough compar- ison of their accuracy and performance, considering the nature of the data. Model accuracy is commonly assessed by measuring the total number of cor- rect predictions relative to all predictions made.As a key performance met- ric for evaluating the efficacy of the models, the R-square statistic is widely employed. Four algorithms have been tested on a selected dataset, and their performance has been compared to identify the most effective model for the given data
)