#Machine learning - a simple Intro

-Abbhirami

Machine learning can be viewed in 2 different ways.
My classification:
1. Learning about the machine's working to get approximate results
2. Machine trying to learn
To talk about the first one, before taking a machine to do something with that. We basically study their a. Hardware
        b. Software and etc
Hardware dealing with it's speed, its capacity to store and etc
Software that deals with the OS, the version, the application. And with this knowledge trying to get some results

The latter one wherein the machine is learning with nothing but only - DATA.
Data which is composed of so many details in it helps the machine to make some prediction. The amount of data that is fed into the system plays a significant role in fetching close to 80% close to the accurate result. 

Now, here is the classification given by Scientists
1. Learning about the machine to fetch particular results - Supervised Learning
2. Machine trying to learn-- Unsupervised Learning

Having understood this basics, let us jump into real world examples:
where does this "machine learning" thing can actually be identified? How can I actually start finding out areas that actually uses "machine learning"?
These are some simple question I find people asking, quite a lot of time.

Simple, you open your mail box and you find a ton of spam mails and immediately you select them and report it spam. Now what happens? Your machine(not the machine literally) identifies the emails and make sure that next time you don't get those.

To actually find out the domains on internet where machine learning is used, try to watch closely the e-commerce website, you search for a product and then you abandon you search. Now you can see similar products being listed or recommended for you, not just on e-commerce, even on YouTube too. This is unsupervised learning. No firm can sit write code for each one of you, trying to find out your preferences. 

Here, the machines are so designed that they actually start studying our pattern of preferences, the tasks you and me are doing(I mean the searches) and gains an experience out of those data. With the learnt data about you and me, it actually refines to give better performances.

Hope you have a little idea about it. Up next will be a post dealing with much more on ML ;)