Alcohol Quality Analysis Using Machine Learning Regression Technique

This research paper investigates the predictive modeling of alcohol quality machine learning regression analysis. The study employs a comprehensive approach to preprocessing, modeling, and evaluation to achieve accurate predictions. Initial data exploration reveals the dataset's structure and identifies potential issues such as missing values and outliers. Preprocessing steps include outlier removal using the interquartile range method and feature scaling to ensure uniformity. The dataset is split into two parts, i.e., training and testing, for development and evaluation. Regression models are built using the ordinary least squares method, incorporating features transformed to address skewness and scaled to improve model performance. Model diagnostics are conducted to assess assumptions including multicollinearity and normality of residuals. The final model exhibits strong predictive capability, as evidenced by high R-squared values and low mean squared error. Visualization techniques illustrate the relationship between actual and predicted values, providing insight into model accuracy. Overall, this paper demonstrates the effectiveness of linear regression in predicting alcohol content with an accuracy of 92%, offering valuable implications for quality control and production optimization in the alcohol industry along with other regression models such as gradient boosting with an accuracy of 90% and decision tree regressor with an accuracy of 83%.