Description
Course info
Rating
(16)
Level
Advanced
Updated
Dec 15, 2020
Duration
1h 5m
Your 10-day individual free trial includes:

Expert-led courses

Keep up with the pace of change with thousands of expert-led, in-depth courses.
Description

Time series predictions are difficult and the rise of neural networks and TensorFlow has made generating highly performant machine learning models possible. In this course, Implement Time Series Analysis, Forecasting, and Prediction with TensorFlow 2.0, you’ll learn how to build models with multiple TensorFlow model types and be able to select the highest performing model. First, you’ll explore time series cross validation and how to create a baseline. Next, you’ll discover how to use neural networks on a single step ahead process. Finally, you’ll learn how to expand the modeling technique to predict multiple time periods in advance along with generating multiple simultaneous predictions on different series. When you’re finished with this course, you’ll have the skills and knowledge of TensorFlow needed to build models for good time series predictions.

About the author
About the author

Chase is currently Lead Data Scientist at Tesorio and formerly was an Assistant Professor of Finance and Economics at the University of South Carolina Upstate.

More from the author
Building Successful Virtual Teams
Beginner
1h 48m
Jun 11, 2020
More courses by Chase DeHan
Section Introduction Transcripts
Section Introduction Transcripts

Course Overview
Hi everyone. My name is Chase DeHan, and welcome my course, Implementing Time Series Analysis, Forecasting, and Prediction with TensorFlow 2.0. I currently lead data science at Tesorio and hold a PhD in Economics from the University of Utah. In this course, we're going to cover the neural network approach to time series forecasting and prediction using TensorFlow. TensorFlow provides alternative ways to generate these time series models that aren't classical, statistical methods. This makes them easier to implement, and in many, but not all cases, able to generate better performing forecasts. Some of the major topics that we will cover include the basics of time series analysis, proper time series cross validation and windowing, and finally, the ability to generate univariate and time series forecasts using covariate methods. By the end of this course, you'll know how to build time series models with a variety of different models and have the foundation to experiment building your own. Before beginning the course, we should be familiar with Python and TensorFlow. This course will use many of the modeling concepts you've learned in the past with a twist of applying the models to time series data. I hope you'll join me on this journey with the Time Series Analysis, Forecasting, and Prediction in TensorFlow course, at Pluralsight.