What Is machine Learning
One of the symptoms of human intelligence is to be able to incorporate learning, and achieve more efficiency over time as new data is integrated into the knowledgebase of your brain.
|What Is machine Learning|
Traditionally, computers are not considered intelligent because they have excellent computational capabilities, they must be programmed to use this power, and there is no ability to integrate learning. Instead, they should be re-programmed to include improvements - often referred to as Program 2.0 so that a better version of the software running on the computer and the better version of the candidate can be nominated.
Machine Learning is a branch of computing that incorporates algorithms for analyzing input data, and can predict on output through statistical analysis, whereas with new data being available, updating predicted output for.
In other words, the algorithm allows the computer to include new data, and over time updates its algorithm, so that learning is being done effectively. A closely related, and sometimes synonyms are considered artificial intelligence - although some would argue that artificial intelligence is considered a broad word in which machine learning is a sub type.
The date for the phrase machine learning is 1959, when Arthur Samuel, a pioneer in computer gaming and artificial intelligence, and a Stanford University research professor defined it as "ability to learn without programming explicitly".
In relation to checkers, he was interested in learning the machine, which he considered to be an ideal subject due to the simplicity of the game. Due to the lack of computing power available at that time, instead of trying to run every possibility, its algorithms have alpha-beta pruning (a version of the minimax algorithm) to select a move based on the position of the pieces, including the location. The pieces of king used, and the possibility of a win
He kept his principles in practice in 1961 when his program won a match against the Connecticut State Checker Champion, at that time, was considered as the fourth ranking player in the country, who used to give credit to his work.
It paved the way for more ground breaking work in the field of machine learning. It includes a famous match of 1997, where IBM Supercomputer Deep Blue defeated world champion Gary Kasparov in the series of matches in a more complex game of chess after the initial loss of years ago.
Recently in 2016, Google took another more complex game of Go, a popular Chinese board game that is known for its high level of strategy. Using AlfaGo Algorithms for machine learning, a product of Google Deepmind, a professional player, was beaten in five consecutive effective games.
Machine learning has been implemented more than just games. In 2012 in Google X Labs, a machine learning algorithm was designed to go through YouTube videos, and freely identify those who have a cat in the video stream. By 2014, Facebook had machine learning algorithms, Deepface, which could match images of the faces of more than 97% accuracy, which reached the performance of a specific human at the time of that work.
In 2015, to facilitate more projects, Microsoft introduced its distributed machine learning toolkit, which currently includes distributed (multi-word) word embedding for high quality natural language processing.
Machine learning techniques have also been implemented on robots and their ability to do complex tasks in an autonomous manner has also been implemented. Has been interested in military applications, and it has given birth to many technical giants including Stephen Hawking and Steve Vojnick to send an open letter to the United Nations.
His concern is that weapon-based machine learning represents "the third revolution in war". On the other hand, autonomous technology promises to make cars safer to drive, and it was recently demonstrated with the technology implemented at Speed's Goodwood Festival in a classic vehicle, 1965 Ford Mustang.
Occupations have also embraced learning machine, and an example is an automated chatbutt that reduces the level of customer interaction with the more expensive customer support staff.
There is also the tendency to go away from the phone interaction menu ("Press one for it, press two for it") which generally disturb customers towards text interaction. For example, at the end of 2017, the Royal Bank of Scotland launched its new AI Chatboot, Luvo, a web chat tool that opens on the bank's website and asks if the customer has any questions or not.
The goal is to help Luvo clients to help 10% customers directly with the ability to answer straightforward questions, to provide more complex support to the right human agent to guide others with more complex issues. The thinking is that by handling the simple tasks of Luo, it frees human advisors for more complex customer issues.
This special virtual chatboot is powered by IBM's Watson Conversation Tool, which is considered to be one of the most widely advanced AI engines with special power in natural language detection. This incorporation of machine learning in the form of Chebbut in the Royal Bank of Scotland is part of a broad trend, driven by cost and time savings for customer service, which has embraced other companies with a clear dividend.
However, Facebook has discontinued its text-based chatbot, M at the beginning of this year, so the transition is rarely limited, because the demand for viable applications is being sought for this technique.
Music for your ears
Machine learning has also been implemented on online music streaming. On Spotify, the on-demand music service that extends with over 100 million users, the firm implements machine learning to stream music that matches your music taste. The 'Discover Weekly' feature of Spottifi is particularly popular, which appears to be a handmade list of songs based on your listening habits - like a good friend who mixes your listening habits with new artists Knows about making tape.
However, this machine is achieved through learning algorithms which include data from collaborative filtering that analyzes your listening behavior and similar fans, as well as natural language processing that includes text, and directly The track's audio is analyzed.
All this data goes through the company's machine learning algorithm to generate these music challenges, and undoubtedly contributes to the popularity of Spotify - while some competing services, such as Sonja and Pandora, Spottifi do not have to Chooses songs based on tagging.
With the ability to learn the machine for the best human champions in many strategic board games, the power of these algorithms which can incorporate new data into the decision making process, is clearly displayed. And now, through autonomous driving, through proprietary driving, diverse industries from customer service, have the power to incorporate machine-learning algorithms.