Haya Al-Kuwari

(back to home)
Houda Bouamor: Machine Learning

Machine Learning is the study of pattern recognition and computational learning theory in artificial intelligence. It studies the construction of algorithms and learns from it to make predictions on the data. It's really similar to computational statistics. Also, it's considered as a subfield of computer science. Arthur Samuel, an American pioneer in the field of computer gaming, artificial intelligence and machine learning, defined it as the "field of study that gives computer the ability to learn without being explicitly programmed".

Machine Learning is believed that it was discovered first in 1763 when Thomas Bayes wrote "An Essay towards solving a Problem in the Doctrine of Chances" and it talked about the theory of probability. Then, in 1950, Alan Turing, a computer scientist, mathematician, logician, cryptanalyst and theoretical biologist, proposes a 'learning machine' that can learn and become artificially intelligent. In 1952, Arthur Samuel joins IBM's Laboratory and starts working on the very first machine learning programs and created programs that can play checkers. In 1992, Gerald Tesauro develops a computer backgammon program and it was able to rival one of the top human backgammon players. In 1998, MNIST database was released. MNIST database was able to evaluate handwriting recognition and a team, led by Yann LeCun, developed it. Also, IBM's Watson, question answering computer system, beat two human champions in Jeopardy competition. Much more achievements were made later by machine learning, showing that it's evolving in a really high speed.

It uses algorithms like neural networks, decision trees, logic search optimization techniques, etc. Also it uses comprehensive data quality and management, interactive data exploration and visualization of model results, comparisons of different machine learning models, GUIs, automated ensemble model evaluation so it identifies the best performers, easy model deployment for reliable, quick and repeatable results, and an integrated end-to-end platform so it pairs the best algorithms.

Machine Learning is important because it would be quickly to automatically produce models that analyzes complex data and delivers it faster and more accurately. Also, an organization would have a better chance of identifying good opportunities while avoiding risks. One of the most famous examples in today's world is a self-driving car; where it uses algorithm to learn and collect data and improves its own driving by time. Also, online sites like Netflix and Amazon uses Machine Learning. More general examples are like financial services (like banks), government, health cares, marketing and sales, oil and gas, and transportation.

Questions about Machine Learning:
1. What is the difference between Data Mining and Machine Learning?
2. How is Artificial Intelligence and Machine Learning are different?
3. How does the machine actually learns from the data and statistics?


(word count; 404) | resources; i ii iii iv |

REVISED VER:

Machine Learning is the study of pattern recognition and computational learning theory in artificial intelligence. It studies the construction of algorithms and learns from it to make predictions on the data. It's really similar to computational statistics. It's also considered a subfield of computer science. Arthur Samuel, an American pioneer in the field of computer gaming, artificial intelligence and machine learning, defined it as the “field of study that gives computers the ability to learn without being explicitly programmed". Machine Learning uses algorithms like neural networks, decision trees, logic search optimization techniques, etc.

Machine Learning is believed to have been discovered first in 1763 when Thomas Bayes wrote “An Essay towards solving a Problem in the Doctrine of Chances" that talked about the theory of probability. Then, in 1950, Alan Turing, a computer scientist, mathematician, logician, proposed a ‘learning machine' that can learn and become artificially intelligent. In 1952, Arthur Samuel joined IBM's Laboratory and started working on the very first machine learning programs and created programs that can play checkers. In 1992, Gerald Tesauro developed a computer backgammon program and it was able to rival one of the top human backgammon players. In 1998, the MNIST database was released. MNIST database was able to evaluate handwriting recognition and a team led by Yann LeCun, developed it. Also, IBM's Watson, a question-answering computer system, beat two human champions in a Jeopardy competition. Many more achievements were made later by machine learning, showing that it's evolving at a really high speed.

Machine Learning is important because it quickly produces models that analyze complex data and delivers it faster and more accurately. Also, an organization would have a better chance of identifying good opportunities while avoiding risks. One of the most famous examples in today's world is a self-driving car that uses algorithms to learn and collect data and improves its own driving by time. Also, online sites like Netflix and Amazon uses Machine Learning. Amazon uses it to build models that predicts and detects fraud and inappropriate item review. More general examples are financial services (like banks), government, health cares, marketing and sales, oil and gas, and transportation.