Machine Learning can be categorized into following categories:
- Supervised Learning
- Unsupervised Learning
- Reinforcement Learning
For example, when you were a kid, solutions were provided to you by your parents. These solutions and decisions were based on their past experience. You learnt as you grew. That is Supervised Learning.
However, when you start taking your own decisions without any such result or outcome-based intelligence, that is Unsupervised Learning. Same is true with the machine.
Let’s have a look at it technically.
Supervised Learning
In case of supervised learning:
- The data is labelled, which means both the dependent and independent variables are provided.
-
There is an Input variable “X” or set of input variables and an output variable “Y”.
Y = f(X)
- The function is approximated to predict new values of Y given X.
-
Examples:
- Regression: Output variable is a real value such as Amount, Height, Weight, etc.
- Classification: Output variable is category, such as Yes, No, red, Blue, Yellow, etc.
| Loan_ID | Gender | Married | Dependents | Self_Employed | Income | LoanAmt | Term | Credit History | Property_Area | Status |
| LP001002 | Male | No | 0 | No | $5849.00 | 60 | 1 | Urban | Y | |
| LP001003 | Male | Yes | 1 | No | $4583.00 | 120 | 1 | Rural | N | |
| LP001005 | Male | Yes | 0 | Yes | $3000.00 | $66.00 | 60 | 1 | Urban | Y |
| LP001006 | Male | Yes | 2 | No | $2583.00 | $120.00 | 60 | 1 | Urban | Y |
Unsupervised Learning
In case of Unsupervised Learning:
- Only X (input) variable is known.
- The goal for Unsupervised Learning is to model the underlying structure or distribution in the data in order to learn more about the data.
- There is no correct answer here and there Is no teacher.
- Algorithms are left on their own to discover and present the interesting structure in the data.
-
Example:
- Clustering: Customer behaviour grouping, such as spending patterns, their preferences, their likes and dislikes.
- Association: Recommendation model, for a product which may be liked by a set of customers.
Reinforcement Learning
How do we know if those decisions taken in Unsupervised manner are going to give us desired result? There is where the Reinforcement Learning help us by rewarding the good decisions and penalizing the bad ones.
The idea is to maximize the gain or reward. It’s something like you pay your bills on time and the result is a good credit history.
Similarly, Supervised Learning may group a set of customers correctly using their spending pattern and let’s say now your marketing department wants to show them a set of advertisements. Now, without any past data, how do we maximize the reward or gain, which in this case Is click through rate and this is achieved by reinforcement learning.
