# A dashboard illustrating bivariate time series forecasting with `ahead`

[This article was first published on

Want to share your content on python-bloggers? click here.

**T. Moudiki's Webpage - Python**, and kindly contributed to python-bloggers]. (You can report issue about the content on this page here)Want to share your content on python-bloggers? click here.

Here is a link to a dashboard illustrating bivariate time series forecasting with the package ahead:

https://thierry.shinyapps.io/ridge2shiny/

This dashboard is more specifically about `ahead::ridge2f`

(in R) and `ahead.Ridge2Regressor`

(in Python) **hyperparameters’ meaning and impact**. In the first two rows of the figure, everything related to `ahead::ridge2f`

and `ahead.Ridge2Regressor`

is colored in blue, in-sample and out-of-sample, whereas input series’ observed values are colored in red. Here are **a few things you could try**:

**Illustrating**Leave every other parameter constant – to their default value. Set the number of lags to 3, and increase the number of nodes in the hidden layer*overfitting*:`nb_hidden`

. Observe what happens on the right (two first rows of the figure), when the input is perfectly fitted.**Illustrating**Leave every other parameter constant – to their default value. Increase \(\lambda_1\), and observe the regression coefficients (third row of the figure) associated to the original features \(x_1, x_2, \ldots\) being shrinked towards zero.*shrinkage*:**Illustrating**Leave every other parameter constant – to their default value. Increase \(\lambda_2\), and observe the regression coefficients (third row of the figure) associated to the hidden layer \(h_1, h_2, \ldots\) being shrinked towards zero.*shrinkage*2:

To

**leave a comment**for the author, please follow the link and comment on their blog:**T. Moudiki's Webpage - Python**.Want to share your content on python-bloggers? click here.