Using A-priori and other techniques to mine for frequent item sets

What you'll learn

Learn about frequent item-sets, their applications in analytics and how to mine for such information in big data.

Description

While working with an ex Fraud Analytics startup Shantanu had the chance to develop parts of a system that made use of Frequent ItemSets to help develop an automatic mined insight generation tool.

Frequent item-sets or FIS as they are also known as, as a concept, establish the building block of many other associated concepts like association rules, “also-bought-with” recommendations etc.

In this talk Shantanu would go into the basics of FIS, and related concepts and talk about how, given a dataset we can use A-priori , FP-growth and other interesting techniques to discover frequent item sets, the related association rules etc.

Apart from theory we would also deal with implementation of FIS systems using Apriori and FP-Growth

Prerequisite knowledge

Knowledge of basic relational algebra/databases would be helpful. Some rudimentary amount of probability would also come in handy.

Speaker Bio:

Shantanu is presently a software consultant at Sahaj.

Shantanu has got a huge body of experience in many areas of software engineering as diverse as embedded systems to rule based expert systems to enterprise Java systems and lately has also worked with analytics applications. That late involvement has also brought about his actual exploration of many areas of data-science which includes data mining. This paper is based on his experience of helping build part of an analytics application.