Julia Programming Projects

Training our data models

Machine learning can be divided into four main types, depending on the methodology and the type of data that is used:

• Supervised
• Unsupervised
• Semi-supervised
• Reinforcement

In supervised learning, we start with a dataset that contains training (or teaching) data, where each record is labeled, representing both input (let's call it X), and output values (named Y). Then, the algorithm's job is to identify a function f from input to output, so that Y = f(X). Once this function is identified, it can be used on new data (that is, new inputs that are not labeled) to predict the output. Depending on the type of output that needs to be computed, if the output has to be assigned to a certain class of entities (as in, it represents ...