The passage to relate to Machine Learning and Deep Learning!

Original article was published on Deep Learning on Medium

The passage to relate to Machine Learning and Deep Learning!

How does the machine learning works:-

Let me explain to you in simple terms. Imagine you are sitting in your room and want to go to a cloth shop which is 10 km away from your house, using your car. There are three straight ways you can go to that shop. The first way has high toll tax, narrow lanes but has nice quality of road; Second way has narrow roads, good quality of road and no toll tax; Third way has no toll tax, no Broadway, and bad quality of the road. Which way you would take?

Clearly you would avoid the third way because it has a bad quality of road and also it does not have any broadway for your car. You would think about using either first way or Second way. Now you might try both the roads for some days regularly and see where you feel much comfortable traveling. That would be a personal choice which one you make.

So what your mind did here? It eliminated the option which was mostly irrelevant to your requirement. Second, it started doing hit and trial and see which of the other roads fit you the best. After repetitive usage, you learned that the second road was best suited for you.

The same thing happens in machines. When we give a machine an output to achieve, it searches for all the ways it can achieve it. Eliminates ways that are irrelevant. Tries all the ways it can achieve output and searches for the best way out. That is Machine Learning.

How does deep learning works:-

These factors would have come into the picture while deciding which road was best for you! First, you will search for the way which has the least number of Red lights or way where you meet maximum green lights. These are called Nodes. Nodes are like milestones, before reaching your output. Next is the weight of the nodes. Weight is analogous to the traffic it has. More traffic it has, lesser the rate of comfort you would have. Every node has weight (but inversely proportional. More weight is better). Third and the final thing is Feedback. Feedback is analogous to the comparison of time you expected you should have taken and the amount of time you actually took. End of the day you would know if the path you took was better than the previous day or not. This is how you can Rate and Rank the ways to get the best way out.

In one short statement, Machine learning is a repetitive process done by the machine in which it rates and ranks all the attempts and takes the better one next time.

Thank you for reading

And I am open to discussion.