By Sheena Goel
Before taking a look at ML’s history, it is important to understand difference between
Simple Computing, AI and ML.
According to the father of Artificial Intelligence, John McCarthy, AI is "The science and
engineering of making intelligent machines, especially intelligent computer
programs. Whereas, machine learning is a branch of computer science in which you
devise or study the design of algorithms that can learn. It is subset of AI and
way of learning AI.
Try google for history of ML and you will find that net is inundated with plethora of
websites all having almost identical information on this subject.
About 70 years from now, in 1952 Arthur Samuel (IBM) wrote a program which could
play checkers. It was called Eliza and helped improving performance of checkers
players. But it was not mature system and served only as a proof of concept.
Perceptron, was a machine designed in 1957 to take inputs like (pixel of images) and
create an output.
A few years later in 1959, “neural network” a system modeled on the human nervous
system used an adaptive filter to remove echoes over phone lines and is still in use today.
As the time passed by, many more researchers contributed towards the growth of ML,
NetTalk was one of them which was created in 1985 to learn to pronounce written English
text, which now has evolved into text to speech.
IBM’s computer Deep Blue developed in 1997 to play chess was capable enough to
evaluate 200 million positions per second. And beat chess grandmaster Garry Kasparov.
Come 2010, Microsoft developed motion sensing input device name Kinect which could
track 20 human characteristics at a rate of 30 times per second and allowed people to
interact with computer via movements and gestures.
Since then ML has come long way and is more advance with neural networks
implementation.
Watson from IBM and Google Brain from Google won US game show Jeopardy in 2011
In the year 2012 an influential research paper was published on ImageNet Classification
and computer vision, describing a model that can dramatically reduce the error rate in
image recognition systems.
Also, Google’s X Lab developed an ML algorithm capable of autonomously browsing
YouTube videos to identify the videos that contain cats.
In 2014, Facebook developed DeepFace an algorithm, which can recognize an dverify
individuals on photos with hight accuracy
By 2015, Amazon launched it’s ML platform and same year Microsoft created distributed ML
Toolkit, to enable efficient distritution of ML problems across multiple computers.
Google’s AlphaGo, won world’s hardest board games “Go” in 2015
Same year, Paypal used ML programs in a hybrid approach where Human detectives
defined the characterstics of criminal behavior and ML programs used those parameters to
identify the bad actors from the site.
Alphabet’s Jigsaw in 2017 learned to identify online trolling by reading millions of website
comments.
Now a days we have Alexa from Amazon, which identifies human voice and the instructions
given by it and does your bidding.There are self-driving cards from Tesla, developed using neural networks and deep learning.
Let’s see what the furture of ML unfolds for the human beings. Which could be as
dangerous as it sounds fascinating.
References:
1. https://www.datacamp.com/community/tutorials/machine-deep-
learning?utm_source=adwords_ppc&utm_campaignid=10267161064&utm_adgroupid=1028423017
92&utm_device=c&utm_keyword=&utm_matchtype=b&utm_network=g&utm_adpostion=&ut
m_creative=332602034364&utm_targetid=aud-392016246653:dsa-
429603003980&utm_loc_interest_ms=&utm_loc_physical_ms=1007820&gclid=EAIaIQobChMIo
tb7jNCS6gIVhoWRCh2CiA5NEAAYAiAAEgLygvD_BwE
2. https://mlplatform.nl/what-is-machine-learning/
3. https://cloud.withgoogle.com/build/data-analytics/explore-history-machine-learning/
4. https://www.forbes.com/sites/bernardmarr/2016/02/19/a-short-history-of-machine-learning-
Comentarios