Machine Learning - Introduction : A new era of intelligences

Hello there, it’s been a while, I’m writing my next post. In this post I am writing about a really interesting topic and now a days one of the most popular one, nothing but Machine Learning.

When I first heard about machine learning, I was just “WOW…!!!!, cooollll”  (O.O) and a picture of robot(from I robot movie) came into my mind and I thought about how amazing it

would be if machine can learn by itself. 

In the end of this post you will know that what machine learning actually is. So, here I start.

Machine learning, it is very popular term in computer science and one of the growing concept of Artificial Intelligence field. Machine learning as name suggest is a process in which machine learns itself, but there is no particular or fix definition of machine learning. If you Google it you find many different definitions are there, from many computer scientists.

But one of the famous definition by Tom Mitchell in his book Machine Learning is

                              “The field of machine learning is concerned 

                           with the question of how to construct 

                           computer programs that automatically 

                           improve with experience.”

                                                                                    -Tom Mitchell

In simple term, machine learning is about learning form examples and experiences rather than from written manual rules.  Machine has no brain or understanding power. It just do what it is programmed for without any understanding. But using machine learning concept it can understand the data and form analysis of that data. It can also predict or make a decision based on analytic data. 

Like if you seen the Captain America: Civil War than you might remember that in the last fight (iron-man vs Captain)F.R.I.D.A.Y.(Iron-man suit’s AI) found  the pattern in Captain’s attack and make a counter attack program. That is what machine learning can do.

Take an example like a program that can recognise an apple from image. A data set that contains the attributes or in terms of machine learning it called features of apple like colour, size, shape etc. are already provided. So when we give an image of apple it will check in dataset if the features value of that apple is equal or not. If it is, then it says that it is apple else no. Now what If we give an image of apple which(apple) has a size which is not available in program dataset or database..?

Here, machine learning concept comes in the picture. First, ML algorithm analyses the data and finds the pattern in that and then it checks the input (here apple’s size). Probability of giving correct answer can be increased more than 80-90% using ML concept. What we need to do is write an algorithm for finding the pattern in dataset instead of writing manual rules. It is very cool isn’t it?

So, now I hope you got an idea about machine learning.

Ok than, that’s all for this post. In the next post we will see the application of machine learning and some references which are useful to jump in to this awesome ocean.  ta-ta bye-bye :)