What are the 3 types of machine learning?, the answer is Supervised machine learning, Unsupervised machine learning, and reinforcement learning. These are the 3 types of machine learning. This article will be a brief overview of all the 3 types. So if you wanted to get a snack start in machine learning this article is for you.
What machine learning is actually all about. In traditional Computing before machine learning came, what we used to do. We used to give some input to the computer and give it some rules, and the computer used to perform the processing based on rules and given us the output. For example, I will Define my role as output = input x 2. I gave the input as 5 and computed the processing and gave me output as 10. But in machine learning, things changed a bit.
Now, we started giving computers the input and output like
- 2 input and output is 4
- 3 is input-output is 6.
and we’ll get him several other rows following this relation and machine learning algorithms are supposed to understand this rule. This is what our computers are getting intelligence. They are doing learning on their own they are able to find out the rule the relationship existing between the input and the output.
Types of machine learning
There are three types of machine learning supervised, unsupervised, and reinforcement learning.
Supervised machine learning
This is where we will be labeling our data set. If we want to classify between cats and dogs. So we will label the images as cat and dog and train it to our machine learning models. Mostly deep nonlinear models and they will understand from the labels that this type of image is a cat and this type of image is a dog.
We won’t have labeled data set. The examples can be something like. When there is a very big digital library and they have thousands of hundreds of books. They want to classify each of the books. So it’s very difficult for a human to go and region every book and classify them. What we will do we will apply unsupervised learning in which the model Will classify the books into different categories. Maybe that category was a book of physics. So they will be classified as one category. Similarly.
We can also have a category for Bio, in a similar category format. So this way unsupervised learning works
Reinforcement learning works in the sense where we don’t even give our model supervised data set or unsupervised data set. We will give the model a scenario like an environment and an agent. It’s just like training a child to walk. So the child will try to walk fall again it will learn from mistakes and correct itself. And finally, he will learn to walk. Similarly, we won’t train a robot how to work. The machine will get some reward if the decision is correct. So based on the d’bari reward and Punishment given for the wrong decisions, the model will understand many scenarios.
In this scenario, which is not possible. Like we can’t do these things using supervised learning or unsupervised learning. So reinforcement learning will have an environment and agent and where the agent will learn many different scenarios. So we have various types of algorithms.
- Face recognization in-camera apps.
- The suggestion of products on e-commerce sites.
- Youtube video suggestions.
- Playlist suggestion on Spotify.
If you wanted to learn machine learning in detail with some online courses you can check How to start machine learning. Best guide. This article has details on how to start a carrier in machine learning.
Are all based on a machine learning model. People pursue machine learning not only for jobs but also for increasing their own businesses. Finally, before leaving make sure you share What are the 3 types of machine learning? Best guide. within your community. All of this will encourage me to make more and more blogs. Please write in your comments in the comments. Thank you very much.