Supervised Machine Learning

In the last post by Atharva Peshkar, you learned about Machine Learning in detail. Well, Machine Learning algorithm can be divided into three parts depending upon the type of dataset, rather say on basis of learning approaches which are Supervised, Unsupervised and Reinforcement Machine Learning.

But in this section, I will be only discussing ‘Supervised Machine Learning’. We will be learning about Unsupervised Machine Learning and Reinforcement Machine Learning in the coming posts.

Most of the practical Machine Learning Programs are based on Supervised Machine Learning. In Supervised Learning, we have the input( say X) and we have the output(say Y) for the same input. And we teach the machine to map the relation between X and Y.

Example: If we have input and output as shown in the following table.

After analysing this input-output computer will be able to draft a relation between X and Y and say   Y=4(X).

Now we can use this computer-generated function to find a value of Y for any value of X. Say for X=2; Y=8 or for X=7; Y=28.

Thus, the data (Input-Output table) given to the program is known as Training data and the data used to test the function is called Testing Data.

In this case, the data was regular. There may be some cases where the data is slightly irregular. Example:

In the above dataset, the values of Y corresponding to X is slightly non-uniform(See the Red Cross). But in order to find a fixed relation between X and Y, Computer will draft such a relation which will only closely differ by the actual relation. This can be illustrated by the blue line drawn on the graph of the dataset. Now the relation between X and Y will be the same as that of the equation of the blue line (Y=mX+C).

Supervised learning can be divided into regression and classification algorithm.

• Classification: A classification problem is when the output variable is a category, such as “Dog” or “Cat” or “disease” and “no disease”.
• Regression: A regression problem is when the output variable is a real value, such as shown in the Housing Price Dataset illustrated above.

We will be learning about these in deep in coming posts. If you have any doubt, query, suggestion or complaint regarding this post. Please reach us out through the Contact Us tab.

Stay Tuned!