Automation of Prediction Method for Supervised Learning

Data science helps to build the prediction model for decision making through data mining techniques. Data Scientists require more time and effort in algorithm selection and data preparation for obtaining more effective results. At the present, a number of existing methods are available for the prediction purpose through supervised learning and unsupervised learning which are based on manual processes. This paper proposed the method for the automation of the prediction process for supervised learning, using prediction type selection methods with the data cleaning process through computing the missing value. It has been illustrated by examples with experimental analysis. This proposed automation methodology makes competitive prediction performance with less effort than the non-automated process.