Predictive or Supervised learning approach Look at the supervised machine learning approach and how to notate it and the theoretical background behind it. The goal here is to learn a mapping from inputs that are called X to outputs that are called Y, given a labeled set of input-output pairs. In supervised machine learning, we used to give data as a PA as the input and its corresponding label. And here, see this is the notation of machine supervised learning where D equals a set of (X_i, Y_i) where i varies from 1 to n. Here, D stands for the training set, which comprises data, n number of data, and its corresponding label. That means, actually, it is written in this notational form. D is the training set, n is the number of training examples. Here, X_i denotes input. This input could be either an n-dimensional vector of numbers. Like, suppose if our data is saying the height and weight of a person, accordingly, we classify them into obese or non-obese, like that. Or it may be a complex structured object. Means, the data could be an image, or it could be, in the case of image classification, or it could be a sentence in the case of sentiment analysis. It could even be an email message, where we use it for spam classification. And it even could be a time series data, like molecular shape, or it may be a graph, etc. And the output is the Y, which is also termed as the response variable. It could be either categorical or nominal variable. If your task is classification, like we can classify or categorize the people as male or female, or anything. Any prediction is true or false, like so classification or pattern recognition may be of categorical or nominal variable output. And in case of regression kind of task, the output could be a real-valued scalar. For example, what may be the income level, or what may be the stock price in the next coming days, like that is what prediction. So, regression tasks use real-valued scalars. Suppose if you are doing any ordinal regression, your Y could be any natural ordering like A to F or X to something, or like any grade, one to five.