Statistical Models vs. Machine Learning: A Comparison in Predictive Analytics

Authors

Krishna Mohan Pitchikala

Abstract

Over the past few years, with the rise of the Internet, the amount of data collected from individuals has increased drastically and is anticipated to reach even greater levels in the years to come. To deal with this extensive amount of information, enhanced methods are developed to process and analyze them. With the latest advancements in technologies like machine learning and big data, we can now handle large datasets, clean them, and use them to make predictions based on past experiences. This process is known as predictive analytics, which helps forecast future outcomes using historical data. In prediction analytics and forecasting the results, two commonly used methods are statistical models and machine learning models. Both are aimed at predicting any events or trends that are likely to occur in the future based on the available records, but there are some differences between the approaches. These differences include how many assumptions about the data they make, the degree of easiness of their results in interpretation, their degree of flexibility in the solution of different problems, and their ability to scale. This paper will analyze how statistical models deviate from machine learning models in terms of their merits and demerits. Additionally, it will suggest the best applicable situations for the above methods and analyze relevant literature to show how the two models are used in prediction.

Keywords

-

Statistical Models vs. Machine Learning: A Comparison in Predictive Analytics. Krishna Mohan Pitchikala. 2020. IJIRCT, Volume 6, Issue 4. Pages 1-5. https://www.ijirct.org/viewPaper.php?paperId=2411081