What is Machine Learning ?

Machine Learning is a subfield of Artificial Intelligence where we train our computer to analyze and learn from the data and make predictions based on the patterns that it has learned.

Artificial Intelligence | Machine Learning | Deep Learning

This is the very famous diagram when talking about AI ( Artificial Intelligence ) , ML ( Machine Learning ) & DL ( Deep Learning) . here in the above image you can see Machine Learning is the subset of Artificial Intelligence .

What is Artificial Intelligence ?

I like to understand or present AI as giving the human intelligence to machines to mimic human behavior or abilities . The goal of AI or the Artificial Intelligence is to create or make machines more intelligent ( probably than humans ) where they can perform task autonomously [ which means the machine would perform the task without the human's intervention ] , make predictions and interact with the environment in a more intelligent manner .

What is Deep Learning ?

As in the above picture, you can see that Deep Learning is the subfield of Machine Learning . It focuses on training neural networks to learn and make predictions or decisions . It is inspired by the structure and function of the human brain by using deep neural networks having multiple layers . Furthermore, it has gained significant attention in various domains like image recognition , speech recognition , NLP ( natural language processing ) and many more.

Traditional Programming v/s Machine Learning

Now , let's talk about What's the limitation of Traditional Programming and why do we actually need Machine Learning ?

Let's understand this with an example : Suppose you are given a task to write a program that calculates the sum of the numbers given to it as an input [ number of inputs can vary ] . First solving the problem using traditional programming ; so as a developer you would create a function that would take let's say 2 numbers and return the sum what if I want to calculate the sum of three numbers, In that case our program would fail . Like may create a function that would like to take N numbers as an input where N is a finite number, but our program would fail in case of N+1 inputs . Now , let's solve this problem using the Machine Learning ; so we can prepare a dataset [ CSV file ] having the different number of inputs and their sum . Means to say , dataset carrying data as 2+3 , 5 , 1+2+3 , 6 , 1+2+9+1+6 , 19 and so on , here our machine learning algorithm would analyze the patterns in the given dataset that it has to find out the sum of all the inputs passed by the user . So here like we have seen a problem statement which cannot be solved using traditional programming but can easily be solved using Machine Learning .

let's take one more example : Suppose you are asked to write a program for self-driving a car, can we really write a program to do so . No! because there are some many cases that we need to explicitly code and the truth is we our self don't know how many cases are there . So here comes Machine Learning to our rescue.

Traditional Programming Approach v/s Machine Learning Approach :

In traditional programming we provide the data and the program/rules as an input, and we get the output ; for ex : to calculate the sum of two numbers we need to provide the numbers num1 & num2 and the program/rule which is num1 + num2 as an input ; based on the input values 2 , 3 we will get 5 as an output .

Whereas in case of Machine Learning things are a bit different . We provide the data and the output as the input, and we get the rules as an output . taking the similar case of calculating the sum of two numbers we need to provide num1 & num2 and the output , if num1 = 2 & num2 = 3 we would provide 2 , 3 and 5 as input . Our algorithm would analyze the pattern in our dataset [ for our Machine Learning model to learn and predict the output effectively we need to provide large number of dataset ] and we would get the rule as the output ; in this case it is the sum of the input numbers as 2 + 3 = 5 .

This is my first blog from 100 days of ML challenge , where I would be writing blogs on topics that I am learning in Machine Learning to explain it to you in simple words .

Please Like this blog if it added some value to you .

Thank-you !!