Скачать или смотреть Advanced Time Series Prediction for Smart Buildings

This video focuses on short-term forecasting of building time-series, comparing four distinct machine learning algorithms: Gaussian processes, gradient boosted trees, neural nets, and recurrent neural nets (RNN). The project details the performance of the models on two distinct target features, each enriched with varying levels of supplementary knowledge, evaluating their impact on predictive accuracy. It's tailored toward building industry professionals interested in the predictive modeling of building data. This knowledge can empower informed, data-driven decisions for increased building performance.

Find more about me @   / blairbirdsell  

How Gradient Boosted Trees Work:    • Advanced Applications of Gradient Boosting...