Why is my MSE cost going up?

[Christian_Simonis]

Hi @monstercameron

I would suggest a residual analysis with the metric that describes your business problem in the best way and working with:

• a training set,
• (a validation set) - in brackets since you seem to have ~25 labels if I interpret your last plot correctly and assumed is it complete concerning your labels
• and a test set (that was never seen by the model before).

I believe here you will find a thread that describes how to do that:

How to evaluate accuracy of a regression model

Excellent! Congrats! As a hint for a potential next step. Also: Cross validation and residual analysis can be useful. Especially the latter can be very helpful since it visualises the whole distribution of your training and e.g validation set and you can easily compare it against your business requirements (e.g. when predicting a pressure in a regression task, your business requirement might be: less than 110 mbar pressure error allowed. ) The residuals should be normally distributed an…

see also this repo.

Hope that helps!

Best regards
Christian