Sharmistha Sikdar’s research focuses on developing statistical and machine-learning methods to solve empirical problems in marketing. Some of the applications of her research methods include predicting customers' multichannel engagement and purchase behavior, and competitive price dynamics on e-commerce platforms.
RESEARCH INTERESTS
Multivariate Modeling Problems, Pricing, Ecommerce, Customer Relationship Management, Customer Analytics, Multichannel Marketing, Statistical Modeling, Machine Learning, Markov Models, Bayesian Methods
PUBLICATIONS
- Sikdar, Sharmistha, Giles Hooker and Vrinda Kadiyali, “Variable Importance Measures for Multivariate Random Forests”, Forthcoming, Journal of Data Science.
MANUSCRIPTS UNDER REVIEW OR REVISION
- Sikdar, Sharmistha, Vrinda Kadiyali and Giles Hooker, “Characterizing Price Dynamics on Amazon Marketplace”
- Sikdar, Sharmistha, Ishita Chakraborty and Nika Dogonadze, “Neither a Picasso nor a da Vinci: A Multi-modal Model for Pricing of Novice Artwork”
- Vana, Prasad, Sharmistha Sikdar and Vrinda Kadiyali, “Does Amazon Have Pricing Power? Evidence from Pricing during the Covid-19 Pandemic”
- Sikdar, Sharmistha and Giles Hooker, “A Hidden Semi-Markov Model of Multi-channel Customer Engagement Dynamics”
RESEARCH IN PROGRESS
- With Wayne Taylor, Nikita Borale, and Scott Neslin, “Creating More Successful Customer Journeys by Managing Holidays and Abandonment”
- With Mansur Khamitov and Andrés Cuneo, “Brand Positioning Mismatch: Effects on Own, Sibling and Parent Brands”
- With Luminita Enache and Felipe Bastos Gurgel Silva, “How do Firms Respond to Population Aging: Evidence from Product Innovation.”