Prediction Using Inferential Statistics Programs

Abstract

The purpose of this Final Degree Project is to delve into a very interesting concept related to obtaining and extracting relevant information that we can find in a data collection. We are talking about Data Mining. Under this name all those techniques that help us extract relevant knowledge and information that are implicit in the databases are grouped. Raw information can be much more useful and easier to work with if it is ordered, classified and divided or grouped into common concepts. These two tasks are addressed by data mining, it provides us with tools that classify and group this "raw" data and thus be able to get the most out of it. However, it is not the only thing we can achieve by applying different data mining methods and techniques. Through these mechanisms of calculation, association and segmentation, through the search for common patterns in the data, situations that are always repeated or implicit "rules" in the data itself, we are able to predict different situations or data that we are going to receive in a future. The classic example for this is the shopping cart. Through simple methods of analyzing purchases made in any supermarket, looking for patterns of behavior and, as we have mentioned, situations that are repeated on several occasions, we can predict that, for example, when someone buys hamburgers, there is a high probability that they will also buy hamburger bun. At first glance it may seem like a somewhat obvious prediction, but it is not so obvious when patterns of behavior of people when buying are discovered and, always by analyzing the data obtained, we reach the conclusion that placing supermarket products in one distribution or another can become very relevant when registering more purchases of some products or others.