<![CDATA[Prediction of student academic achievement is a very important research area; this can be seen from the many researchers who conduct research in this area. To make predictions, a machine learning model is needed. Along with their parameters, the majority of machine learning models have associated hyperparameters. However, knowing the right mix of hyperparameters is essential for robust model performance. A methodical procedure called hyperparameter optimization (HPO) aids in determining the appropriate values for them. In this study we compared four hyperparameters tuning techniques, namely HyperOpt, Random Search, Optuna and Grid Search. The results of the hyperparameters from each of these techniques are then used in machine learning algorithms to predict student academic achievement. Validation uses the 5-fold cross validation method while performance testing uses Mean absolute error. From the experimental results it was found that the hyperparameter technique The best method for predicting student academic achievement in machine learning models is gridsearchcv.]]>
Copyrights © 2024
