Supervised Learning

Basically, supervised learning can be separated by classification and regression. Nowadays, supervised learning is a key thing many problems can be solved by. but it still needs tons of data so that can get ready. meaning need tons of data is supervised learning learns by data to solve the problem. so that is named by 'supervised'. 

 

We normally split data into training, test and validation. Everyone wants to have a highly accurate model for their problems. So after we set the model and test the model using the test data and we finally measure the model' accuracy by validation data. so it allows the model more accurate. 

 

A role of regression is to predict what would happen based on what has happened before such as prediction of the price of a house, the price of flight ticket so forth. for me, it is amazing and interesting as I love to imagine about the future. There are various ways of regressions such as Linear Regression, Logistic Regression, Polynomial Regression etc.. (we will talk about them next time). Before we use regression in order to build a model. we need to consider how this algorithm would affect our problem and help our work. Because it is not a perfect solution in every case. 

 

Linear regression                                                                                            classification