Decision Science in Action

I joined   CoreCompete   on  Feb 18, 2013 Monday. It has been a great learning and exciting experience. When I walked into Banjara Hills, Hyderabad office of Core Compete (now shifted to Gachibowli, 3rd office) only thought in my mind was did I take wise decision by choosing Core Compete over a global giant.  Today, I can say that it was the best professional decision I have taken. I feel proud and amazed when I think about our growth from 4 to 200+ in period of 6 years in the niche business of advance analytics and big data technology and recognized as fastest growing consulting organisation in North America ( http://www.consultingmag.com/fastest-growing-firms-2016/   ) for last three consecutive years. Yes, we are growing fast, but that’s not the only reason that I love working at Core Compete.  In last 6 years I worked for multiple clients across multiple domain using different advance analytical techniques and technology platform.I got opportun...

Product propensity models are developed to identify   the  customers or observations those  have high likelihood of responding to any cross-sell campaign or any event of interest.  Model d evelopment which includes has multiple steps requires extensive efforts. 1. Identification of business objective or business problem. 2. Identifying analytical objective  3. Identifying data scope and time window.  4. Data preparation and exploration 5. Data treatment and variable reduction 6. Model training and validation 7. Model selection and interpretation  8. Business approval  Steps to automate the model scoring for its periodical usage. 9. Model deployment  10. Periodical Model Scoring  Once the models are scored and used in the campaign for targeting customers to cross sell the product of interest, effectiveness of model must be measured. I am going to discuss the different ways to ev...