In this chapter, we looked at the theory behind the logistic regression model and how it solves the binary classification problem. Binary classification is a fundamental problem. You can naturally extend logistic regression to multiple class classification problems.
Ultimately, you learned about three ways in which you can implement a logistic regression model on the web. First, we looked at the TensorFlow.js Core API, which is suitable if we want to implement the algorithm in any way we like. It is capable of covering any kind of use case that can be solved by the operation graph. Then, we looked at the Layers API, which is useful if we want to construct a model that has a simple stack of neural layers. This API can help us build logistic regression but also proves its merit when it's used to create deep learning applications. Finally, we introduced machinelearn...