Answer for Why Validation loss is always less than previous loss?

https://365datascience.com/dwqa-answer/answer-for-why-validation-loss-is-always-less-than-previous-loss/ -

Hi Keerthi,

The reason for that is that once the validation loss stops increasing, the training is over. Therefore, the validation loss is always decreasing during training. So great observation!

Best,

Iliya

#365datascience #DataScience #data #science #365datascience #BigData #tutorial #infographic #career #salary #education #howto #scientist #engineer #course #engineer #MachineLearning #machine #learning #certificate #udemy