In this chapter, we covered some of the common and best practices that are used in solving machine learning or deep learning problems. We covered various important steps such as creating problem statements, choosing the algorithm, beating the baseline score, increasing the capacity of the model until it overfits the dataset, applying regularization techniques that can prevent overfitting, increasing the generalization capacity, tuning different parameters of the model or algorithms, and exploring different learning strategies that can be used to train deep learning models optimally and faster.
In the next chapter, we will cover different components that are responsible for building state-of-the-art Convolutional Neural Networks (CNNs). We will also cover transfer learning, which helps us to train image classifiers when little data is available. We will also cover techniques...