NBA Player Score Prediction Based on Machine Learning

: With the success of machine learning and data visualization in many fields, the NBA(National Basketball Association) has also benefited from its huge demand for data analysis. These analysis results have been extensively applied in player draft, player training and tactical decisions, playing a crucial role in management and coaching staff decisions. This article utilizes data visualization technology and machine learning to analyze the NBA dataset. Using random forest and multiple linear regression models to predict NBA player scoring performance, and evaluate the model using R-square scores and MAE(Mean Absolute Error). There are some significant relationships between Points and several features like Turnovers, FGM and Minutes Played. After a ten-fold validation experiment, it was found that both the multiple linear regression and random forest are greater than 0.98 in R-square scores. And according to the result of the comparison, the multiple linear regression model is more suitable as a score prediction model and has a better stability for this dataset.