Model Overfitting

[Aaryan]

Hi All!

I was recently making my own AI/ML project with GCP products and heard about over-fitting! As a result, I was wondering, what are the signs that your model is overfitting, and how can you address them?

[jaia · Accepted solution]

Hello,

Thank you for contacting the Google Cloud Community.

Here I am mentioning some signs of Overfitting:

1. High training accuracy, low validation/test accuracy
2. A model with too many parameters or layers can be more prone to overfitting.
3. Small dataset: With limited data, the model might memorize the training examples instead of learning underlying patterns.
4. Noise in the data: Noise can introduce spurious patterns that the model might learn, leading to overfitting.

Regards,
Jai Ade

[jaia · Accepted solution]

Hello,

Thank you for contacting the Google Cloud Community.

Here I am mentioning some signs of Overfitting:

1. High training accuracy, low validation/test accuracy
2. A model with too many parameters or layers can be more prone to overfitting.
3. Small dataset: With limited data, the model might memorize the training examples instead of learning underlying patterns.
4. Noise in the data: Noise can introduce spurious patterns that the model might learn, leading to overfitting.

Regards,
Jai Ade

[Aaryan · Accepted solution]

Thank you so much! That Answers my question!