Tuesday, 6 March 2018

Google machine learning


Today we are here to learn about google machine learning. Google machine learning is a valuable tool to help engineers at Microsoft, Google and everywhere else make sense of the data. As we already know that, how much demand of the engineers those are using machine learning now a day.

google machine learning



Introduction to google machine learning


Every software engineers, needs to better three things here.
  1. Reduce the time you spend on programming
  2. Customize your product for a specific group of people
  3. Complete "Un-programmable tasks"

Reduce the time you spend on programming


For example I want to write a program to correct spellings, to do so I could do a lot of things and apply some rules of thumbs in such a manner that "e" before "j" and "u" before "o". And after a long week of hard work we may or may not be able to do these. On the other hand we can do this using machine learning tool / google machine learning to feed it some example and get a more reliable program in a small fraction of time.

Customize your product for a specific group of people


Use google machine learning to scale your product. For example I produce my English spelling character writing code by hand, and I got success to do that. Next what if I want to do this for one hundred most popular languages?
The answer is I would start from scratch to do all of these each languages and we take years of effort. But if I built this using google machine learning than it's not a big deal to move towards another language.


How to achieve this using google machine leaning


Just need to collect data in that language and feed that data into the google machine learning model. Remember this model is same as we have already used the data in our first approximation machine learning process of first language.


Complete "Un-programmable tasks"


Google Machine learning let you solve problem that as a programmer you have no idea how to solve it by hand. As a human being I am the ability to recognize human faces and understand their speech. But I do all of these subconsciously. If you ask me to write down the complete code of recognize human being, than I would be completely stuck. But using google machine learning these tasks are performed very well.


How?


I don’t need to tell the algorithm what to do or how to do. I only need to show the algorithm lots of examples and from that the tasks can be solved using machine learning.

Conclusion of google machine learning


Besides these three practical reasons for mastering machine learning there are some more reason that google machine learning change the way of thinking. In machine learning the focus shifts from the mathematical science to a natural science. We are making observations of un-certain world. We are running experiments and using statistics not logic to solve problems and analyzing results. The ability to think like a scientist will expand your horizons and open up new areas that you could not explore without machine learning. 
So for we learnt the introduction of machine learning, in my next article we will discuss about the basic framework which is called(supervised) machine learning.

1 comment: