Lab: Forecast with Time Series

Lab: Forecast with Time Series

Use Cases: Unemployment, Revenue, and Store Sales Forecasting

About this Mission

On this mission, you will build on the concepts from the Time Series Starter. Over a series of activities, you'll build more accurate models with DataRobot using Forecast Distance, Feature Derivation Windows, indicator variables, Known In Advance and Remove from Derivation options, and time series evaluation tools. By building and iteratively improving these models, you will gain the skills to develop your own forecasting models.

Mission format and duration: self-paced, hands-on, 1 hour

Upon completion of this mission, you will be able to: 

  • Build multivariate time series models for multiple use cases
  • Compare models using the Accuracy Over Time (AoT), Stability, and Forecasting Accuracy tools
  • Mitigate the effects of time series regime changes using indicator features
  • Use calendar files to account for events that affect a model
  • Build and evaluate multiseries models to forecast sales across multiple stores

Who should complete this mission?

  • Business Analysts
  • Citizen Data Scientists
  • Data Scientists

Before embarking on this mission, you should complete one of the following:

We recommend that you also complete the following before beginning this mission:

Technical requirements

  • Chrome browser
  • DataRobot Automated Time Series — If you don’t have access to the application, please sign up for our free trial: datarobot.com/trial.

 

Action Items1 hr

  • Activity: Forecasting Unemployment
  • Activity: Dealing with Regime Changes
  • Activity: Forecasting Sales Across Multiple Stores
  • What's Next...
  • Give Us Your Feedback!
  • Resources
  • Community Discussions on Time Series
  • Forecasting Unemployment Datasets
  • Dealing with Regime Changes Datasets
  • Forecasting Sales Across Stores

About this Mission

On this mission, you will build on the concepts from the Time Series Starter. Over a series of activities, you'll build more accurate models with DataRobot using Forecast Distance, Feature Derivation Windows, indicator variables, Known In Advance and Remove from Derivation options, and time series evaluation tools. By building and iteratively improving these models, you will gain the skills to develop your own forecasting models.

Mission format and duration: self-paced, hands-on, 1 hour

Upon completion of this mission, you will be able to: 

  • Build multivariate time series models for multiple use cases
  • Compare models using the Accuracy Over Time (AoT), Stability, and Forecasting Accuracy tools
  • Mitigate the effects of time series regime changes using indicator features
  • Use calendar files to account for events that affect a model
  • Build and evaluate multiseries models to forecast sales across multiple stores

Who should complete this mission?

  • Business Analysts
  • Citizen Data Scientists
  • Data Scientists

Before embarking on this mission, you should complete one of the following:

We recommend that you also complete the following before beginning this mission:

Technical requirements

  • Chrome browser
  • DataRobot Automated Time Series — If you don’t have access to the application, please sign up for our free trial: datarobot.com/trial.

 

Action Items1 hr

  • Activity: Forecasting Unemployment
  • Activity: Dealing with Regime Changes
  • Activity: Forecasting Sales Across Multiple Stores
  • What's Next...
  • Give Us Your Feedback!
  • Resources
  • Community Discussions on Time Series
  • Forecasting Unemployment Datasets
  • Dealing with Regime Changes Datasets
  • Forecasting Sales Across Stores