Data Science: Framework & Methodology
Main Article Content
Abstract
Data science has evolved rapidly in recent years, providing organizations meaningful insight from available huge amount of data. Due to complexity there is a need for comprehensive framework for data science process which will guide practitioners to work efficiently. This paper presents framework for data science which includes environment, problem statement, data preprocessing, data gathering, exploratory data analysis, feature engineering, feature selection, feature extraction, model training, evaluation and deployment. Additionally abstract emphasizes on supervised and unsupervised algorithms required for model selection. The proposed framework will help to tackle with the challenges during data science process. It is a systematic approach which will enable researchers to take decision based on data driven insight. By adopting this framework researcher can streamline data science projects and take accurate decisions.