What is Predictive Analytics?
According to Wikipedia, predictive analytics is an area that encompasses a variety of techniques from statistics, modeling, machine learning, and data mining that analyze current and historical facts to make predictions about future, or otherwise unknown, events.
Predictive analytics enables companies to gain actionable and forward-looking insight from their data.
Predictive analytical applications have already made inroads in Banking, Oil & Gas, Insurance, Healthcare & Retail industries.
Technology Factors That Have Helped Push Predictive Analytical Processing
- Faster CPUs
- Commodity-based massively parallel processing capabilities combined with faster and multi-core CPUs have significantly reduced the time it takes to perform complex analytical processes
- In-Memory Analytics
- Lower memory prices and 64-bit addressing have enabled loading of massive amounts of data directly into memory. Data can be mined mush faster in the memory directly.
- Big Data:
- Technologies like Hadoop, R, MapReduce, and natural processing and text analytics have enabled companies to collect, analyse and mine massive amounts of structured and unstructured data
Predictive Analytics: Applications for an Enterprise
Predictive analytics is much more than general forecasting that most of the companies currently do. Below are some of the key areas where I believe predictive analytical capabilities can be used to improve a company’s competitive edge:
- Revenue Management: Pricing Optimization
- Supply Chain: Inventory Optimization, Safety-stock management, Product-mix planning
- Human Resource: Critical talent retention
- Quality/ R&D: Yield Predictions
- IT Risk Management: Spam filtering, Data-Loss Prevention, Advanced Persistent Threats
Predictive modeling is the process by which a model is created or chosen to try to best predict the probability of an outcome.
A generic modelling process includes collecting data from various source systems, both internal and external. Next step involves creating statistical models based on certain assumptions/hypothesis. These models are then tested using sample data points and eventually a model is selected for deployment. This model is thereafter allowed to mature over a period of time using various game theory and machine learning techniques.
Following modelling techniques currently exist:
- Group Method of data handling
- Naïve Bayes
- k-nearest neighbour algorithm
- Majority Classifier
- Support vector machines
Some of the commonly available and popular open-source solutions include the following:
Predictive Analytics is not a new concept, but recent technological innovations have now enabled us to start using these capabilities faster and cost-effectively. A predictive analytical solution must be an integrated system covering end-to-end processes in a company’s supply chain, customer relationship or financial management processes. Predictive analytical solutions can help uncover hidden opportunities for a company, avoid pricing erosion or lost sales.
Albert Einstein once said: “The definition of insanity is doing the same thing over and over again, while expecting different results.” Investing in a good predictive analytics solution is the best investment a company can make for their business.