- Why do we need machine learning ?
- Examples of machine learning
- How is machine learning different from traditional programming ?
Machine learning is a subdomain of Artificial intelligence and Computer Science . It is the study of ways of using data and mathematical algorithms to mimic human intelligence. In simpler words ‘using data to answer questions‘.
Why do we need machine learning ?
As we(humans) have experienced in our life, the requirement of learning and adapting to changes around us, the same way there is a need to learn and adapt DATA across the Scientific world. Millions of videos in youtube, billions of google searches , billions of data people post online needs to be managed and flown in a smooth way. Not only that, if you look around you , you’ll find many things which are carried out by machines without any human supervision, like weather prediction, spam detection, face detection in mobile phones(read this for more in-depth intuition) etc. and by time this will only tend to grow(exponentially) . Ofcourse we human can’t perceive that amount of data, thats why we need machine learning . And even if you are not aware of it, machine learning is all around us.
Examples of machine learning
ok, lets see some applications of machine learning you are familiar with. When you are using youtube, you probably get feeds at the home page which are similar to what you often watch right ? even though you never subscribed those channels. well yes, the reason we are able to do this is machine learning. when you are watching videos, youtube is monitoring things like the watch time of the kind of videos you are watching, the tags, the kind of videos where you click the like or dislike button and so on. and using this data, they try to predict, what kind of videos you might like, and show it in the home page. thats very similar to what netflix or amazon do, when they are recommending you a movie. even your phone recognizing your face is an example.
How is machine learning different from traditional programming ?
well basically, the theoretical approach of ML is like.
NOTE: program here means algorithm. and then you use the program you got, to get the output of other inputs