Scalable Causal Inference for Down-stream Impact of an Event
Dr. Vineet Chaoji, Manager, Applied Science at Amazon, Bengaluru, India.
Downstream Impact (DSI) measures the longer term impact of a customer action in terms of increase in revenue or units sold. The DSI of an event is critical to making a wide range of business decisions — e.g., how should one price the Kindle Paperwhite, which Prime benefit should be recommended to a customer, which product advertisement should shown to a customer, etc. One can think of millions of events within the Amazon ecosystem. Estimating the DSI of an event is a causal inference problem. The talk will introduce the DSI estimation problem and its connection to causal inference. Few techniques for computing the DSI will be discussed. Finally, I will present a scalable system to estimate the DSI for a large number of events.
Vineet Chaoji is an Applied Science Manager within the Machine Learning team at Amazon where he leads projects related to econometric models of customer behavior, customer targeting and malware detection.
Prior to joining Amazon, he was a Scientist at Yahoo! Labs in Bangalore where his research focused on online advertising and social networks.
Vineet obtained his PhD in Computer Science from Rensselaer Polytechnic Institute.
He has published at top-tier data mining and database conferences and journals. Vineet has also served on the program committees of leading data and web mining conferences.