In simple words Predictive Analytics deals with prediction of future probabilities and trends utilizing statistics modeling and data mining techniques on the historical data.
Think how helpful would it be if, using predictive algorithms and models, if we can identify patterns and predict likelihood of let’s say, possibility of fraud in the credit applications, requirement of frozen tuna cans right after oil spill in gulf, predicting potential number of auto insurance claims in Boulder even before a snow storm hit etc.
Predictive analytics exploits the historical data along with business rules and uses statistical techniques and data mining to predict risk and opportunities in businesses. Learning from the collective experience of your company’s records and producing a predictive score. Predictive model can not only predict outcomes but also the range and probability of the likely outcomes.
Lets have a quick look on what goes inside Predictive Analytics:
It is an iterative process, where predictions are made and weather predictions were successful were not, there is a lesson learnt from this and is used further to improve the process. Statistical techniques like regression, time series models, neural networks etc. have been used actively in this process.
Some of the vendors currently in the market providing some kind of predictive analytics piece with their offering are SAS, IBM’s SPSS, SAP, BO, IBI, Oracle Corp, Kxen Inc., Potrait software, TIBCO Spotfire, FICO, Angoss software etc.