Predicting New York City Taxi Demand - Spatio-Temporal Time Series Forecasting

Explore spatio-temporal time series forecasting in this 40-minute conference talk from MLCon. Dive into a demonstration project predicting taxi demand in Manhattan, NYC for the next hour. Learn basic principles of time series forecasting and compare different models suited for spatio-temporal use cases. Examine the principles of long short-term memory networks and temporal convolutional networks. Discover how these advanced models can decrease prediction error by 40% compared to simple baseline models. Gain insights into making decisions about the future using machine learning and statistics in the field of spatio-temporal forecasting, where predictions encompass both temporal and regional dimensions.

Predicting New York City Taxi demand: spatio-temporal Time Series Forecasting | Fabian Hertwig