Getting started with Machine Learning - Part 2
Hi.. Hello... and welcome back, As this the continuation of the previous article , lets pickup from where we left off.. In previous we saw what is a Classification algorithm. In classification problem we had a set of outcomes which are quantifiable, for example: we had 2 outcomes in case of email spam filter (either the email is spam and or not spam). these outcomes are called " Classes ". So, an input dataset will contain "n" number of unique classes, and the output of the classifier always falls in any one of these unique class. Simply we can say classifier algorithm is like choosing a best answer in a multiple choice question. Here the question is Q. whether a arriving Email is a spam or not? And the choices("Classes") will be. a. "spam" b. "not spam" The classifier has to select best answer based on the Experience(aka. learning or training), it got from looking into thousands - millions of emails. This is all about Classification. ...