Time-series analysis and forecasting is one of the most common and important tasks in business analytics today. The massive production of data through the digitalization of Healthcare, Finance and other industries, the rise of Smart Cities and IoT use cases, will further drive the need for time series analysis.
Time series forecasting problems differ from typical classification or regression problems. They require careful treatment to deal with the time sensitive nature of the data and other practical modelling constraints. In this session, we will share how to utilize time series to predict future values. We will showcase the H2O time-series functionality through practical use cases, and share best practices for running time series experiments with H2O.
Join the session and you will learn:
Traditional time series forecasting solutions and their limitations
Practical considerations in time series forecasting
Time series forecasting using H2O and best practices
Real-world applications including prediction of the COVID-19 impact and other use cases
Dr Vishal Sharma, Senior Data Scientist, H2O.ai
Prior to H2O.ai, Vishal worked in different domains such as finance, smart city and patent analytics. He has lead teams of data scientists and machine learning researchers delivering end-end innovative solutions for lead generation, analytics based corporate lending, machine translation, weather forecasting, regime detection for trading etc.
Vishal has a PhD from National University of Singapore in neural networks and time series. His areas of interest include time series and NLP.