Predictive Analytics is a form of Advanced Analytics but with the special task to make predictions. One of the more common forms of Predictive Analytics is Machine Learning.
Let´s focus on Predictive Analytics from here on, let’s start with a definition by Eric Siegel author of the book ”Predictive analytics: the power to predict who will click, buy, lie or die.”
“Predictive Analytics (PA) – Technology that learns from experience (data) to predict the future behaviour of individuals* in order to drive better decisions.”
* In this definition, individuals, is a broad term that can refer to people (customers, debtors, applicants, employees, students, patients, voters, etc) as well as other organizational elements such as products, locations, vehicles, stores, flights, movies, transactions, etc.
This type of analytics has been through a tremendous development when it comes to methodologies, applications and performance the last five years. The most common mathematical approach is called Machine Learning. It is a branch of Artificial Intelligence and holds a number of algorithms such as Decision Trees, Support Vector Machine and Neural Networks. You encounter various applications of machine learning every day no matter if it is “Siri” in your iPhone, the recommendations done by Amazon and Netflix or ordering an Uber.
Other classic examples are Churn, Loyalty Program, Recommendation and Association Rules analyses.
Optilon offers four main application areas all connected to supply chain;
In addition, there are several other problems where Optilon’s approach can be applied, such as; Association Rules (Hierarchical Basket Analysis, Assortment Optimization and Bundle Generation), Store Clustering, Loyalty Program, Churn Analysis and Internet of Things (e.g condition based maintenance).
The applications we use and recommend for this type of analytics are based on machine learning. They hold a number of algorithms addressing both so called Unsupervised and Supervised problems in a completely new way combining the four key characteristics described below in a very efficient way.
Finally, the most important question you need to ask yourself before taking any predictive analytics initiative is:
“What are the decisions and actions taken by the organization in response to each prediction?”