Stay up to date with our news, ideas and updates.
While developing a deep learning model, we need to train the model on a set of data so that it can learn the patterns within the data. However, we need to ensure that it's not overfitting and generalizes well to new, unseen data.
We all have learned how to build machine learning models using the readily available opensource datasets. But, if you are working on a custom project and the dataset is not readily available, then how you will solve the problem using machine learning algorithms?
In my earlier post, Creating image dataset for Transfer Learning applications in Deep Learning, we have seen how to download the images from websites to prepare custom dataset.
