Summary

In this chapter, we covered transfer learning and leveraged it to create deep learning models faster. We then moved on to learn the importance of separate training, development, and test datasets, followed by a section on dealing with real-life, unprocessed datasets. After that, we talk about what AutoML is and how we can find the most optimal network with little to no work. We learned how to visualize neural network models and training logs.

Now that you have completed this chapter, you are now capable of handling any kind of data to create machine learning models.

Finally, having completed this book, you should now have a strong understanding of the concepts of data science, and should be able to use the Python language to work with different datasets to solve business-case problems. The different concepts that you have learned, including those of preprocessing, data visualization, image augmentation, and human language processing, should have helped in providing you with an overall...