Dr. Ram Gopal talks about the relevance of analytics in the context of decision making – both, learning from past data (predictive analytics) as well as to act in a given context based on trends (prescriptive analytics).
Predictive analytics has two models; patent discovery and predictive models. Patent discovery is called unsupervised learning, whereas predictive models are called supervised learning. In the case of the former, there’s no specific objective to serve but it suddenly appears as a pattern that can be used for predictions. In the case of the latter, there’s a particular objective which is served by the analysis.
Heard the beer-diaper correlation story?
Click here to read the transcript of the talk.
Click here to listen to the transcript of the talk.