Artificial intelligence

Artificial intelligence is a branch of computer science, where human intelligence and thinking power are tried to emulate computers. Artificial intelligence has now become an academic education field where taught how to create computers and software that will show intelligence.

Implementing the technique of human intelligence and thinking through artificial means depends on machine-based artificial intelligence. [1] The computer is brought to the Mimics Cognitive unit so that the computer can think like a human. Such as learning and solving problems. Artificial Intelligence (AI) is a machine-made intelligence. In the field of computer science, the AI ​​field defines itself as a study of "intelligent agents": Any device that can understand its environment and take steps that lead to its achievements far beyond achieving some goals. The term "artificial intelligence" is applied when a machine that functions in "cognitive" functions matches the minds of other people, such as "learning to accept" and "problem solving". Andreyar Kaplan and Michael Henelin have defined the artificial intelligence as "It is the ability to accurately explain the external information of a system, learning from such information and using those lessons to focus specifically on the flexible adaptation."

Once the machines are increasingly able, it is necessary to remove the intelligence from the definition for mental benefits. For example, optical character recognition is not perceived as an "artificial intelligence" example, it becomes a routine technology. The capabilities that have been classified today can successfully understand human speech, strategic game systems (such as chess and going) can participate in higher level competitions, automatically drive, military simulations, and explain complex data.

Ai research can be divided into several sub-sections, which focus on the specific problems, views, use of special tools or the satisfaction of certain applications.

History:

The artificial man who was able to think, originally appeared as a storytelling instrument, could actually start with Raman Lol (1300 AD), perhaps the idea of ​​trying to make an instrument for effective reasoning. With his Calculus RetiCinator, the Gottfried Libyans expanded the concept of mathematical machines (Willemham Skikard did the first engineering work at around 1623), instead of the number, for the purpose of operations operation on concepts. Since the nineteenth century, artificial people became commonplace in the field of scientific fiction, such as Mary Charlie's Frankenstein or Karel Kepek's R.R. (Universal Robots of Rasos) can be mentioned.

The mechanical or "formal" argument study started with philosophers and mathematicians in the past. Mathematical Logic Studies initiated the theory of mathematics of Allan Turing, which can be mathematical by deciding with a machine, "0" and "1" symbols. Through this insight the digital computer could simulate any mechanism of formal logic that was known as Church-Turing Thesis. Neurology, information theory, and the discovery of cybernetics gave the researchers the possibility of building an electrical brain. The traditional design, which is now fully recognized as Ai, was a complete "artificial neuron" for McClull and Pitt's 1943 touring.

AI research was first established in 1956 in a workshop in Dartmouth College. Alan Newell (CMU), Harbert Simon (CMU), John McCarthy (MIT), Marvin Minsky (MIT) and Arthur Samuel (IBM) became the founder and leader of AII research. The newspaper and the students who created the program described him as "astonishing": Computer checkers were victorious, solve problems in algebra, prove logical theories, and able to speak English. In the mid-1960s, funding was funded by the Defense Department for research in the United States and laboratories worldwide. The founders of AI are optimistic about the future: Harbert Simon predicted, "The machine will be able to do what a person can do in twenty years." Mervyn Minkski agreed, "In a generation ... the problem of making artificial intelligence will be solved"

They failed to understand the difficulties of some of the remaining works. Progress was slow, and in 1974, in response to Sir James Lighthal's criticism, the AIA's experimental research was discontinued due to ongoing pressure from the British government and the Congress of the United States. It will be called "AI Winter" in the next few years when the AI ​​project funding was difficult.

Ai research expert was revived by the commercial success of the system in the early 1980's, a form of AI program that imitates human expert knowledge and analytical skills. By 1985, the Ai Bazar had reached more than a billion dollars. At the same time, Japan's fifth-generation computer project inspired the US and the British government to bring back funding for academic research. However, at the beginning of the fall machine market collapse in 1987, AI fell again in corruption and the second chronic recession fell in the state.

In the early 1990s and early 21st century, began to use Ai for supplies, data mining, medical diagnosis and other areas. The success was due to increase in computational power (see Moore's law), solution to specific problems, new relationship between AI and other fields, and the mathematical method of the researchers and the scientific value is of greater importance on a commitment. Deep Blue became the first computer-controlled chess player to defeat a chess champion Gary Kasparov on June 11, 1997.

Improved statistical techniques (known as LITTLE deep learning), access to large amounts of data and rapid progress in computer learning and understanding of computers. By the middle of 2010, machine learning applications were used all over the world. A danger! In answer to the questionnaire of IBM in the Quiz Show exhibition match, Watson defeated two greatest champions Brad Rudder and K Jeanings by a significant margin. Kinet, which provides 3D body-motion interface for the Xbox 360 and Xbox One, which uses algorithms that arise from long Ai research, as a smart personal assistant on smartphones. In March 2016, Alfago became the first computer go-system to beat a professional go-go player without handicaps, winning 4 out of 5 games in a match with champion Lee Seedol. FIFA World Cup 2017 Alfago won three games with Joe, who has achieved world number one ranking for two consecutive years.

According to Bloomberg's Jack Clarke, in 2015, there was a milestone for artificial intelligence, the number of software projects used in Google to use Ai in 2012 has increased in "spideridi usage" in over 2700 projects. Clerk also reported that the error rate for image processing has dropped significantly since 2011. As a result of the rise of the cloud computing infrastructure and the rise of research equipment and datasets, it emphasizes the growth of affordable neural networks. Other examples include Microsoft's Skype system development which can automatically translate from one language to another and the Facebook system can illustrate the image to blind people.

Goal:

The goal of artificial intelligence is to create technology through which computers and machines will be able to work in intelligent ways. Intelligence production (or made) common problems have been divided into several sub-problems. Expect that the special features or capabilities that the researchers would show an intelligent system. The following descriptions received the maximum attention.

Arich Sandwoy emphasizes planning and learning that is relevant and applicable to the given situation.

Logic and problem solving:

Primary researchers have developed algorithms that are step-by-step, as people use them for problem solving or logical cutting. In the late 1980s and early 1990s, AI research was developed to determine uncertain or incomplete information, possibilities and ideas from the economy.

Algorithms may require a lot of computational resources for difficult issues - the most "experience to blast" blurring: the amount of memory or computer time required to solve a specific size problem. Searching for more efficient problem-solving algorithms is getting a lot of priority.

Instead of stepping down by using primarily quick, customizable conclusions, primary AI research has given the model a look. Ai has progressed using the "sub-symbolic" problem solution: the posing agent emphasizes the efficiency of the sensormotor from the higher argument; The pursuit of postgraduate research imitates the structure of the brain; Ai's main goal is to emulate human ability.