Machine Learning , How Exactly Does it Work?

Source

Machine learning is one of the best parts of Artificial intelligence but many don't know what machine learning is and how does it work?

The Machine learning process begins with the insertion of data (training data) into a selected algorithm, that can be utilized for solving the given problem. The data is known or unknown data to develop the final machine learning algorithm. New input data is tested in the algorithm and checks whether it works correctly. The prediction results are checked against the training data. If the prediction doesn't match then the algorithm is retrained multiple times until we get the desired output.

EXAMPLE:
Suppose we want to train a computer in the game "chess". It normally would not be able to make smart moves and beat a human but with ml it is possible!
So basically, we feed the rules of chess into the computer (Data) then we make the computer fight itself multiple times. After a few thousand computers vs Computer chess games with the algorithm, it learns what moves are needed when a certain move is made in order to win. The best part is a computer can do this thousands of times and still work at 'max efficiency'. This way at one point it will beat humans in a game of chess.

It does give us a lot of Doctor strange vibes from Infinity war too, where he checked every possible situation in which they would survive. Maybe magic too has machine learning embedded in it :^)
This is an example of machine learning which in other words is very similar to Patten recognition. This method can be applied in any field and be greatly benefited by it.

Source

Different types of Machine Learning:

✔️Supervised Learning
Here we use known or labeled data for the training data (as input for the algorithm). Since the data is known it is therefore known as Supervised learning.
Example:
We pass in data of 50 oranges after training it will definitely be able to identify whether or not it's orange.
Some best algorithms under supervised learning are
-> Linear regression
->Polynomial regression
->Logistic regression

✔️ Unsupervised Learning
Here the data is unknown or unlabelled in the sense its new data. This data is passed into a machine learning algorithm and is used to train a model.The trained model will try to differentiate and make a pattern out of the unknown data provided.
Example:
Here we insert data of various fruits like oranges, bandanas, custard Apple etc. it would be able to differentiate among them after the model is trained.
Some of the best algorithms under unsupervised learning are:
->Singular Value decomposition
->K-meand clustering
->Partial least squares

Source

✔️ Reinforced Learning
Here the algorithm discovers data through a process of trial and error then decides which results are more accurate.
The Major components of reinforced learning are
->agent
->environment
->action

Hope you learned something new from this post! Please let me know your thoughts on machine learning in the comments.
Thank you for your time ^^