Day 8
You are training your Deep Learning model. After training, you realize the training accuracy is good. However, validation accuracy is poor. What could have been the reason behind it?
Avenir Light is a clean and stylish font favored by designers. It's easy on the eyes and a great go to font for titles, paragraphs & more.
It might be due to overfitting or the validation samples might be of different class.
If your model is overly complex, tries to incorporate minor details, it tends to bad generalization, leading to overfitting which in turn is the indication of high training performance. However, low validation performance.
On the other hand, sometimes a train test split would have been such that the validation data contained the data from rare class or outliers or anything which wasn't in resemblance to the train set. In such a case, you would prefer to use Cross Validation approach so as to obtain an overall average metric for your model performance.