Evaluation of Applications for Nursery School using Machine Learning Approach

Abstract

This Paper demonstrates how several classification techniques can be used to perform classify nursery dataset. There are various classification techniques that can be used to classify dataset. Therefore, the need to choose the right classification technique can affect the result of classification accuracy of the model. Naïve Bayes, Support Vector Machine and k-Nearest Neighbours are three machine learning models that used in this study to evaluate the application of nursery schools. This paper is aimed to find the best machine learning models that can be used to classify the nursery dataset. The pre-processing tasks that have been conducted in this study include data cleaning, attribute or feature reduction, feature engineering technique and dimensional reduction. The data that has been undergoing pre-processed was then used in the third stage to identify the best machine learning technique by develop a comparative analysis between the three chosen techniques. Outcome of this study demonstrates the accuracy of each classification techniques and the accuracy result of three classification models were compared before and after tuning parameter was conducted to determine which classification technique is the best to classify the nursery dataset. The impact of the study can be used as a reference to develop classification model for this kind dataset.