Machine Learning - A Biased Formula or An Art

Time for a post on Machine Learning has been too late, as a proper detailed idea on Machine Learning is never a stabled one.

Obviously, this article too won't be stable for readers.

Yes, Machine Learning is so rapid on its development - but the reality is statistics of firms slowing replace them with computers.

I personally believe," I don't do something without any need " and also nothing is free here.

Before defining Machine Learning - we must know what made this need. If you look back with doing business.

In the 1990's we got introduced to computers into business management as a data collection machine, it controlled the maths around the business.

The Record notebooks have been slowly replaced with Hard disks.

I am not going against Machine Learning, but I am not surprised by it too.

Literally, All the Logical and also probable decisions are taken by computers now with the help of Formulae.

The Risk factor of theory to real-world becomes a major issue and all these are trying to bridge the gap with more and more data and its processing that has been done with either Machine Learning or deep learning model.

What we are trying with the help of Machine Learning is always surprises me, we trying to predict stocks, product recommendations, future of things, future of life and more.

But, it is like a nutshell we revolve around the data in hand, out of way thinking, new trials, and new scope with respect to humanity is slowly eroding.

An Art evolves with time, but data will never

Even a few people tried capturing human thoughts as data for deep learning model, we missed a huge bus, we came here with lots of trial and error evolutions, as we might found errors on this and move on soon.

Process of Humans are three stages for being a Machine, 

• Collecting human data
• Observing it 24x7
• making decisions

To be a Machine, we need to do it ever.

Google says it, " Solve Anything " with Data

But, it is a mechanical problem, not a human thing - we are here to live not for solving.

We unearth a problem and we solve it. because Maths is just made for an easy lifestyle, not for living. 

For example, to solve a problem either politically or business way - all these machine learning algorithms will eventually be biased based on the required result.

From this, I can easily get to the famous quote,

" What you need !  is your Right" 

here we created data and we messed it and now we are reusing it to create a thought impulse than nothing new or useful.

All these data and decisions are helping to sell an unwanted thing to your customers in a possible way.

If sales become my need, then I will bias the algorithm to look like it. The commercialization becomes and ideology rather than what human need.

For example, the national Horticulture board collected crop rates to solve the issue of pricing for Agriculture in India. But, they missed few things and also data becoming a false resulted in the unavailability of  Onions in a few months back. 

Here, mistakes depend on several things like wrong data, errors, a poor examination of data and more, but if these mistakes are made by biasing the algorithm for using it.

"Art can be a Math but An Art can never be Biased."