What Is Machine Learning ????

Lots of People thinks ML is very fascinating word and actually it is very interesting Field. people thinks it is very complex area of study which contains huge programming and mathematics.
Without this we won’t able to get started with ML. but in this article, I’m going to share information about Machine learning from beginning.

Term machine learning is coined by Arthur Samuel in 1959. It is very old term in actual but now days it is growing faster and faster because of its multiple applications. everyone wants to go into the field of machine learning. ML typically comes under the field of AI. AI is very vast field itself which is again part of Data science.

What is ML ?

Have you ever seen any small kid, how he explore the world around him, how they are trying to learn new things. in the beginning they faces difficulty but slowly they become better with experience. This same thing happens with ML also.
Consider ML as small kid. We needed to teach lots of things to this kid in the form of programming. there is concept of model we mostly used in machine learning. This model is nothing but program.
There are various types of models are available in ML. we will see that in our upcoming article.
In short machine learning means, this is technology in which machine is learning from its past experience. Basically it is used for prediction purpose. In order to predict the output or unknown values ML is used

Difference between Traditional Programming and Machine Learning

Traditional Programming:

In this kind of programming we are giving input and some rules to program and it provides us output

Machine Learning Model:

In this we are providing both input and output to machine and machine provides us rules.

Types Of ML:

1. Supervised Machine Learning

In this type of machine learning we are giving both input and output to the machine. consider this algorithm as teacher or supervisor which teach students about everything step by step.
Input is also known as labeled data and output as unlabeled data.

There are two categories of supervised learning
a) Regression
b) Classification

we will see that in depth in our upcoming articles

consider This example: Classification of Iron Man and Captain America
If you need to classify between this both then you need to give features which distinguish iron man and Captain America and you need to provide output which is iron man and which is captain America.
Features are like property of any object which provides uniqueness to that object.
In above example also we are using features, consider following features for above case Color of Suit, type of weapon they are using ,action type, facial type
In this case along with features we are providing output also that’s why it is known as Supervised Learning

For all the cases in ML we are going to use Dataset

2. Unsupervised Machine Learning:

In this type of ML we are just providing Input to the Machine. In this case there is no any teacher for machine, machine needs to find the output itself. there is no any labeled data used here.
There are Two categories of Unsupervised ML
a)Clustering
b)Association
We will see that in later

Consider another Example: Clustering of Spiders and Bats
Clustering means grouping of same category of objects. In above example we need to do separate grouping of Spiders and Justice Bats. Model is try to find the pattern among the object and according to that they forms cluster. whenever there new object comes this model will try to find pattern between them and according to that model will send that object into particular group of cluster.

3. Reinforcement Learning:

In this type of machine learning concept of agents are used. here there is no any kind of supervisor is present. we are not providing any input and output to machine. It totally based on Action-Reward concept.

Thanks For Reading..!

Stay Tuned For Next Article..