To keep the code complexity to a minimum, we will be extracting all the helper methods that we would be recursively calling, as seen in the previous example. We can start with the train() method because that is going to be called first to determine the root decision node.
Before we do that, let's create an Injectable service for our ID3 algorithm in the utils folder which we will be injecting where we wish to use it. This logic can live anywhere you wish, server or client side. One thing to note is that the dataset, in this case, is relatively small, so training the dataset and predicting the outcomes are an okay thing to do on the client side. With larger datasets, which take much longer to train, it is recommended ...
