Data Science Seminar Date: Friday, 10 November 2023 Time: 2.00 PM Venue: NKN Hall A Suite of AI Applications for measuring Advertising Effectiveness Dinesh Krithivasan Kantar. 10-11-23 Abstract Can we leverage the latest advancements in artificial intelligence and machine learning to learn what makes a good advertisement? Kantar has a vast repository of hundreds of thousands of advertisements along with insights on how well they have performed historically. In this talk, we will showcase a wide variety of AI techniques that can be used on this dataset to answer questions like "How well can we predict an ad's performance?", "What makes an ad effective?", "Which ads transfer well across markets?" etc. To answer these questions, we use techniques ranging from Bayesian matrix factorization to deep neural networks and Generative AI. About the speaker: Dinesh is a Partner at the Data Science & Innovation team at Kantar Analytics where he has been leading several R&D initiatives on developing state-of-the-art solutions powered by advanced data science techniques. He has extensive experience in using deep learning and machine learning algorithms across a variety of problems such as computer vision, time-series modeling, text analysis and ROI modeling. He has directed studies for multiple clients across different geographies and industries, addressing issues such as segmentation and targeting, customer satisfaction and loyalty, and brand strategy. Dinesh received his Ph.D. from the University of Michigan, Ann Arbor.
|