Hierarchical forecasting is a unique time series problem with several challenges because of time series dependencies.
An example is Walmart Store Sales from the M5 Competition.
š Walmart has stores in each state.
š Each state has multiple stores.
š Each store has multiple departments.
š Each department has categories of items.
š Each category of items has 100s or 1000s of items.
š Walmart has stores in each state.
š Each state has multiple stores.
š Each store has multiple departments.
š Each department has categories of items.
š Each category of items has 100s or 1000s of items.
This makes for a large-scale time series problem...
And if you can solve this problem, you become valuable to your (future) organization.
And if you can solve this problem, you become valuable to your (future) organization.
How can I help?
I'm working on a new project called my Lost Time Series Modules, which extends my High-Performance Time Series Course.
And the 5th lost module covers how to do hierarchical forecasting leveraging the top 20 solutions to the M5 Competition.
I'm working on a new project called my Lost Time Series Modules, which extends my High-Performance Time Series Course.
And the 5th lost module covers how to do hierarchical forecasting leveraging the top 20 solutions to the M5 Competition.
The lost time series modules are insanely valuable.
5+ hours of expanded time series coding lessons on common yet advanced time series problems.
This includes:
5+ hours of expanded time series coding lessons on common yet advanced time series problems.
This includes:
1. Changing demand
Forecasting in recessions and covid when the demand has changed and you have little information to process forecasts from
Forecasting in recessions and covid when the demand has changed and you have little information to process forecasts from
2. Iterative forecasting
Exposing the new modeltime API for iterative modeling at scale
Exposing the new modeltime API for iterative modeling at scale
3. Recursive + Iterative Forecasting
Extending iterative forecasting workflow for regressors and using autoregression with short-term lags [lags < forecast horizon] to eke out additional performance
Extending iterative forecasting workflow for regressors and using autoregression with short-term lags [lags < forecast horizon] to eke out additional performance
4. Automatic Forecasting
Discover how to use H2O AutoML to forecast with 100's of models at scale
Discover how to use H2O AutoML to forecast with 100's of models at scale
5. Hierarchical forecasting
Tackle the large-scale forecasting challenge that billion dollar Retail, Energy, and Marketing companies face
Tackle the large-scale forecasting challenge that billion dollar Retail, Energy, and Marketing companies face
In total, you'll get:
1. 1563 lines of advanced time series code
2. 5 modules on advanced time series concepts
3. And 6 hours of expert guidance...
... in just the lost time series modules course.
1. 1563 lines of advanced time series code
2. 5 modules on advanced time series concepts
3. And 6 hours of expert guidance...
... in just the lost time series modules course.
If you are interested in learning more, I'm giving away the pre-launch Lost Time Series Modules as part of my 5-Course R-Track during our Christmas Sale.
Christmas Sale Ends Friday.
Christmas Sale Ends Friday.
***Christmas Sale: 5-Course R-Track + The Lost Time Series Module***
Save money and get my new lost time series modules for free when you purchase my 5-Course R-Track Half OFF.
learn.business-science.io/christmas
Save money and get my new lost time series modules for free when you purchase my 5-Course R-Track Half OFF.
learn.business-science.io/christmas
ā¢ ā¢ ā¢
Missing some Tweet in this thread? You can try to
force a refresh
ć
