Добавил:
Upload Опубликованный материал нарушает ваши авторские права? Сообщите нам.
Вуз: Предмет: Файл:
Engish_for_IT_Students.doc
Скачиваний:
74
Добавлен:
25.11.2019
Размер:
838.14 Кб
Скачать

Some Examples of ai Techniques

Artificial intelligence may be applied in many different ways depending on the problem to be solved and the resources available. Some common techniques include the following:

Decision Trees. These software guides are simply maps that tell the computer what to do next based on each decision it makes. Each decision leads to a new branch with new decisions and consequences.

Rules-Based Systems. These systems work by following a set of rules given by the programmer. So long as the programmer has anticipated every possible circumstance that the program may encounter, it can solve any problem.

Feedback. This technique is used to modify programs. Basically, a feedback system monitors the results of a solution to see if the solution worked or in what areas it failed.

Knowledge-Based Systems. These systems are similar to a rules-based system, but they use feedback to learn from their mistakes. As a result, knowledge-based systems can actually learn to solve new problems.

Heuristics. This software technique is something like a recipe for a problem-solving approach rather than an algorithm that solves a specific problem.

Building an Artificial Brain

To create a true artificial intelligence, scientists could try building an artificial brain called a neural network. The human brain consists of billions and even trillions of neurons, each with as many as a million connections to other neurons. Scientists have identified hundreds of different types of neurons and more than fifty different patterns of neuron connections. This level of complexity is simply beyond any computer currently in existence. Even the most powerful parallel computers with tens of thousands of processors don't come close to equaling the number or variety of connections in a human brain.

Unit 7. Networks and the Internet Networking at Home and Around the World Data Communications Over Standard Telephone Lines

Networks (especially the Internet and large WANs, with nodes spread over a large geographic area) commonly transmit data across telephone lines. Although telephone companies are offering more digital lines (which are better suited to data transmission), most homes and businesses are still served by analog telephone lines. To transfer digital data over analog telephone lines, computers must use modems. When a computer sends data, its modem translates digital data into analog signals for transmission over standard telephone lines. At the receiving end, the computer's modem converts the analog signals back into digital data. The most important factors to consider when choosing a modem are internal versus external, transmission speed, data compression, and error correction. Modem transmission speeds are measured in bits per second (bps). Currently, the preferred standards for modems are 56.6 Kbps and higher.

Using digital connections, business networks and homes can transmit data many times faster than is possible over standard telephone lines. In areas where digital connection is not possible, homes and businesses are connected with standard analog lines, but high-speed digital lines are run between the telephone company's switching stations. The most popular digital telephone services are integrated services digital network (ISDN), T1, T3, and DSL. They offer faster data transfer rates and higher bandwidths than standard telephone lines.

Соседние файлы в предмете [НЕСОРТИРОВАННОЕ]