Performance Analysis of Machine Learning-Based Scheduling Policies for Operating Systems

This study investigates the effectiveness of machine learning-based approaches for predicting a target variable in a given dataset. We compare the performance of several popular regression models, including linear regression, random forest regression, gradient boosting regression, support vector regression, and neural network regression, on a real-world dataset with a large number of features. For each model, several hyperparameters are tuned by using a grid search approach to find the configuration that achieves the best performance on a held-out test set. The obtained results show that the neural network regression model achieved the best test RMSE of 0.173, followed by the gradient boosting regression and random forest regression models with test RMSEs of around 0.18. The linear regression and support vector regression models had higher test RMSEs, indicating lower effectiveness in this particular dataset. The research findings suggest that more complex machine learning models with a higher number of hyperparameters to tune can perform better for this type of prediction task, but careful hyperparameter tuning is required to achieve the optimal performance.