Use of Deep Learning Models in Estimating Conversion Probabilities and the Contribution of Communication Channels in Advertising Campaigns

Explore a 47-minute talk on how deep learning models can revolutionize marketing attribution strategies. Learn about neural network-based models, particularly those using LSTM architecture, that help marketing specialists predict conversion rates and identify the most effective communication channels for advertising campaigns. The presentation addresses the challenge of budget optimization across multiple advertising channels by integrating historical customer data with attention mechanisms. Discover the proposed Encoder-Decoder framework modification to the original architecture that enhances the ability to make data-driven business decisions about channel relevance and marketing resource allocation.