Abstract
As businesses look to boost flexibility and cut costs, cloud computing is becoming more and more popular. There is always room for improvement, despite the fact that the major cloud service providers use a pay-as-you-go pricing model and provide consumers rapid scaling up and down. CPU and memory utilisation, which represents workload, frequently varies, which causes businesses to incur extra costs and has an adverse effect on the environment. To lessen the impact of this, a Random Forest algorithm which is a supervised machine learning model is proposed for resource scheduling in cloud computing environment. This proposed model is an ensemble algorithm for regression, classification and task decisions that deal with developing decision trees. The proposed strategy is compared to existing machine learning algorithms including XGBoost, Ridge and Lasso. Experimental results show that the proposed algorithm performs better than the compared algorithms in terms of the prediction accuracy of the CPU and memory. Experimental results conducted on go ogle colaboratory using Materna Dataset shows that the proposed algorithm achieved a prediction higher accuracy of 0.1 - 45.1 % higher than the compared algorithms in terms of the R-squared (R 2 ), the Root Mean Square Error (RMSE) and lower percentage errors for Mean Absolute Percentage Error (MAPE).