Predictive analytics techniques and approaches can be broadly categorized into two groups:
- • Regression techniques, and
- • Machine learning techniques.
Regression techniques behold the foundation of predictive analytics techniques. These techniques derive a mathematical equation. The equation derived can be used as a model to represent relationship between all the features or attributes under consideration. Depending upon the application, there are a plethora of techniques that lie under this category. Some of the techniques are:
- • Linear Regression Model, and
- • Logistic Regression etc.
Machine Learning Techniques
Machine Learning, comes under the field of artificial intelligence, came into existence to help computers to learn and understand. Today, it consists of several techniques which find application in wide variety of fields such as fraud detection, medical diagnosis, analysis of stock market, and weather forecasting. Sometimes it becomes very easy to predict the value of unknown dependent variable without analysing the relationship between other variables which are known and independent. However, in some cases, where it is very complex to deduce relationships between variables, machine learning techniques come to the rescue. Some of them are given below:
- • Back Propagation Network,
- • Support Vector Machine, and
- • Naive Bayes, etc.