Over fitting and under fitting

I have just started learning neural network and I am bit confused about this.
How to know if our model is over fitted with data?
How to know if our model is under fitted?

Overfitting is when your model perform well on training data but have a poor performance in actual use or unseen data. Underfitting is the opposite of that.

To answer the question about “How to know”. You simply have to validate/evaluate it by using the model against unseen data and compare the result with the result from training data.

Use this link for your reference

1 Like

Thank you for your reply. That was helpful.