6.5. Overfitting in Machine Learning | Causes for Overfitting and its Prevention

[@Kamalsai369]

Excellent Explanation

[@Ary2345]

less data is the cause of under fitting because our data does get enough experience to learn from it . For over fitting our data set is quite big tha it learns it each amd every feature trend and pattern and apply yhe exactly same trend on new or training data. That's what i learned.

[@guillermogonzalez3088]

Hi. What I don´t understand is how do you define noise in the training dataset? is always outliers or is it something else?

[@sachinvithubone4278]

Awaiting for this video, thank you so much. Overfitting concept is clear.😊 1.As you said training data gets good predict and test data didn't get good prediction.. if we have less data so how can add more data? 2. Or using early stoping or bias variance tradeoff?

[@Siddhardhan]

you are most welcome 😇 low accuracy need not be caused due to less data all the time. there can be other reasons as well. the model couldn't find the patterns in the data. that's one situation. I'll make separate video bias variance trade-off

[@sachinvithubone4278]

@@Siddhardhan thanks

[@jerseyk7238]

@@Siddhardhan Can you please make a video on Bias Variance tradeoff

[@debashisjana1271]

Please start a course on deep learning, neural network.. It will be very useful for us..

[@laharipuneeth6632]

Good one 👌

[@Siddhardhan]

Thanks 😇