Microsoft Research AI Fireside Chats, 2018 (30 mins). A discussion with Eric Horvitz (Chief Scientist at Microsoft) on ride-sharing, mapping, building data science & machine learning teams, and more.
The Cube, Women in Data Science Conference, 2018 (15 mins). An interview about data science and machine learning technologies and career paths.
International Symposium on Forecasting, June 2021 (31 mins). A discussion of travel time (ETA) prediction as it’s used in ride-sharing, including methods used to solve this problem.
National Academies of Sciences, Engineering, and Medicine Webinar, April 2020 (20 mins). A survey of geospatial (map) technologies used in ride-sharing and in emergency vehicle fleet management.
Stastistical and Applied Mathematical Sciences Institute, Dec. 2020 (28 mins). Presents novel theoretical results on the accuracy of various methods for travel time (ETA) prediction.
Technical Vision Talk, Women in Data Science Conference, Stanford University, March 2018 (21 mins). Pricing and matching techniques used in ride-sharing, and their underlying economic models.
Matching & Dynamic Pricing in Ride-Hailing Platforms, MSR AI Distinguished Lectures, Microsoft Research Lab, May 2018 (1 hour). A more in-depth discussion of pricing and matching techniques used in ride-sharing, and the underlying economic and optimization methods.
Model-Based Clustering for Online Crisis Identification in Distributed Computing, Statistics Department, Harvard University, Feb. 2011.
Convergence Rate of Markov Chain Methods for Genomic Motif Discovery, Workshop on Advances in Monte Carlo (“Adap’ski”), Utah, Jan. 2011.
Lower Bounds on the Convergence Time of Adaptive MCMC Methods, International Conference on Monte Carlo and Quasi-Monte Carlo Methods in Scientific Computing, Aug. 2010.
Spatial Inference of Nitrate Concentrations in Groundwater. Cornell Univ. School of Operations Research and Information Engineering Colloquium, Sept. 2008.
Physician Performance Assessment / Bayesian Computation for Multimodal Posterior Distributions. Cornell Univ. School of Operations Research and Information Engineering Colloquium, Feb. 2008.