Introduction

Predictive modeling is probably one of the most exciting fields in data analytics. It has gained a lot of attention in recent years due to massive amounts of data being available in many different verticals. It is very commonly used in areas concerning data mining to forecast future trends.

Predictive modeling is an analysis technique that is used to predict the future behavior of a system. It is a collection of algorithms that can identify the relationship between independent input variables and the target responses. We create a mathematical model, based on observations, and then use this model to estimate what's going to happen in the future.

In predictive modeling, we need to collect data with known responses to train our model. Once we create this model, we validate it using some metrics, and then use it to predict future values. We can use many different types of algorithms to create a predictive model. In this chapter, we will use SVMs to build linear and nonlinear models.

A predictive model is built using a number of features that are likely to influence the behavior of the system. For example, to estimate weather conditions, we may use various types of data, such as temperature, barometric pressure, precipitation, and other atmospheric processes. Similarly, when we deal with other types of systems, we need to decide what factors are likely to influence its behavior and include them as part of the feature vector before training our model.