Predictive Analytics & Big Data
What It Is
Predictive Analytics (PA) is an important leading-edge technology that is being adopted by Fortune 500 corporations and coveted by many other entities in industry and academia. As the name suggests, it seeks to predict the outcome of certain events. The applications of PA are quite wide-ranging. Predictive Analytics provides actionable solutions to many business problems. It’s an extension of web analytics and data mining.
Another term used in the industry for Predictive Analytics is ‘Big Data’. The roots of ‘Big Data’ come from ‘Decision Support System’ that became popular in 1980s. As the name suggests, Big Data uses corporate data to predicts outcomes which aid the decision making process.
Predictive Analytics has been used successfully in many industries, including Health Care, Banking & Finance, Public Utilities, Telecommunications, Media, Retail, Insurance, and Travel. Amongst the major users of PA techniques are UPS, United Health Care, Macys, Bank of America, Citi Group, LinkedIn, Netflix, Amazon, Best Buy, and Proctor & Gamble.
How It Works
PA is based on techniques such as Decision Trees, Naïve Bayesian statistics, Linear and Logical regression, Neural Networks and TreeNet. For unsupervised learning, it also uses correlation and clustering.
Why It Matters
A well-known Predictive Analytics success story is the role it played in predicting the performance of the individual baseball players who propelled the Boston Red Sox to three World Series titles. The Hollywood movie, Moneyball, which is based on a true story, was built on a similar theme. Amongst the more obvious uses:
- Predicting fraud in credit card usage.
- Determining whether a professional athlete’s compensation should be based on predictions of future performance.
- Predicting election outcomes.