Evolution and History Of Machine Learning

From Turing Machine and Arthur Samuel to current highly intelligent robots, How much machine learning has come far! Let’s look at History Of Machine Learning.

Let’s find out through History of Machine Learning.

Before we start, here is the definition of Machine Learning, according to N. Mitchell

“A computer program is said to learn from experience E with respect to some class of tasks T and performance measure P, if its performance at tasks in T, as measured by P, improves with experience E.”

or simply, here in machine learning algorithms, such algorithms are made which can improve their output every time with experience/data passed through algorithm or model.

Now coming to History of Machine Learning:

There are many dots to start the history of Machine Learning. And to start with, we will pick the invention of Computer i.e. 1950 by Alan Turing.

• Alan Turing-1950

  

Alan Turing, a mathematician who initiated Artificial Intelligence during the 1940s and created a “Turing Test” to find whether a computer is intelligent or not?

• Computer Learning Program by Arthur Samuel- 1952
  Arthur Samuel in collaboration with IBM created a machine learning program which learned to play the game Checkers. By the mid-1970’s, this program was capable of beating the human players.

• The Perceptron-1957

  

  Perceptron is nothing but a part of neural networks which was developed by Frank Rosenblatt at the US Office of Naval research for visual recognition.

In 1969, Marine and Saima published a book Perceptron which shows the limitation of perception and neural networks.

• The Nearest Neighbour Algorithm-1967
  The nearest neighbor algorithm allows recognizing the basic pattern in the given data set.

• The Stanford cart-1979Stanford cart, invented by the students of MIT, was capable to navigate obstacles in a room on its own.

• EBL- Explanation Based Learning-1981
  The program analyses the training information and avoids irrelevant information to form a general rule to follow. Interesting!

• Data-Driven Approach to Machine Learning-1990
  Scientists started making applications/programme for a large amount of data. Here, it picks up!

• IBM’s Deep Blue-1997 

In 1997, IBM’s Deep Blue beats the world champion Garry Kasparov at a game of Chess.

• IBM Watson-2011
  Using Natural Language Processing NLP, IBM Watson beats the human champion in the game of Jeopardy.

More to come yet!

A conclusion to History of Machine Learning:

Machine Learning is developing from since long, then why it has become so popular now? The answer lies in big data and cloud storage. Big data helps machine learning in the same way as a catalyst does for a chemical reaction. With the onset of Big Data research, we can say that history is yet to be written.