Saturday 6 June 2020

What is the theory behind Predicting Machines ?

Artificial Intelligence is no longer an alien topic for anyone these days and why would it be if it's  kinda intriguing and most importantly it is in the Economic Market right now, Isn't it? Let's not state the obvious fact here because we are in the age of innovation and machines where every day is advancing with dynamic changes in tech. 

A.I. is coming in to replacing the traditional statistical techniques and tools for doing Prediction. If we look at the business side instead of Comp. Science, A.I. is becoming an ART more than a SCIENCE. It means humans are converting non prediction problems into prediction problems to better understand the machines and put more intelligence into them.

Ages ago, scientists have made a statement that we would not have an autonomous car on the road because there were so many If's and we could not control machines in an uncontrollable environment UNTIL Machine Learning guys realized that instead of making so many predictions, how about making only one prediction and that was - What would humans do ? This was the bang on point where we started acknowledging autonomous cars rolling down the streets efficiently because A.I. helped in making good predictions with time and more learning. 

Let's make a comparison between a human brain and Machines's art of predicting in case of driving. 
Imagine a car, you put an human on the driving seat and A.I. on the passenger seat. Human is driving, processing the data in his monkey brain and driving for miles, taking actions ( right or left), applying breaks, accelerating and going straight. There are so many If''s and When's spinning on his mind like If a man comes in front of the car, he needs to stop or if the light is green, he needs to move and so on.

Reversing the situation, If the A.I. is driving with the help of it's sensory inputs such as camera, radars, lights and other programs, there are so many IF'S and very small no. of When's because A.I. is making only one prediction here which is What would humans do if there is a speed breaker, a sign board for no u-turn, or obeying the traffic lights etc. 
A.I. will observe, take action and will make mistakes too but after they see a pattern of how humans are driving, they make sharp patterns & predictions in their minds and update their model accordingly. 

So in the next few drives, A.I. will be observed as more efficient drivers than Humans because we can make mistakes being humans, getting drunk, dizzy and careless while driving but the A.I. will still predict their problems, making lesser no. of mistakes and by that, their confidence window will grow wider. 

Similar observation has been seen in the fields of Translations and HR domains as well where A.I. has converted non predicting problems into predicting problems to make better use of Cheap Prediction and hence, they are coming up with splendid results. 

At the end, you can see A.I.working on solving a prediction problem and with this we all are taking advantage of this better, cheap and faster technology. 

MyShiksha has offered a platform where industry experts are mentoring students in various domains and their career paths so that whichever path they wanna continue with, they will master it first ! Be it A.I., Digital Marketing, Advertising, Finance etc. 

Stay tuned for further learning !


1 comment:

  1. A well-researched and compelling read. The author's passion for the subject shines through. Thoroughly enjoyed it.
    building toys for kids


Career scope in Digital Marketing

Many of us enroll ourselves into Digital Marketing courses thinking that may be it will help in Career change, better growth opportunities...