I have a lot of discussions with people who work in the tech world and sometimes it can be tricky understanding the differences in some of the disciplines, which causes people to use terms interchangeably. The disciplines have their own unique attributes so I thought I’d write a blog defining some of these disciplines more specifically, just so you know.

The field of predictive analytics uses (for the most part) a combination of three disciplines, namely machine learning, data mining and statistics. Predictive analytics uses these disciplines to build a model to look into the future and predict predict future outcomes for business purposes. Predictive analytics looks forward through time to add value the the business.

Here’s a short summary of the disciplines:


Statistics is a discipline that involves collecting, analyzing, and interpreting data. There is no preprocessing of data in statistics, that’s where data mining comes into the picture. The fundamental difference between statistics and data mining is that in statistics there is no preprocessing of the data. With this in mind, the data used in statistics should be carefully selected.

Data Mining

The goal of data mining is to look for patterns and knowledge in large volumes of data. The goal is not simply to ‘mine’ for the data, but actually for for patterns within it. Data miners can use machine learning, but don’t necessarily have to. Data miners use data classification to label data into a specific category.

Machine Learning

Machine learning involves the construction of algorithms to allow machines to learn and make predictions in data. As the system analyzes more data, it learns more about it and what to expect from it. As new behaviors are processed, the system can make more informed decisions.


Those are some of the distinctions between the disciplines, and should give you some clarity on what they involve. Often the simple things can be overlooked so it always useful to know key differences in various disciplines - especially in the ever evolving world of AI. I’ll be sure to include more definitions in future blogs.