User:Kent lippincott/Artificial neural network

Article Evaluation:

Lead Section:

Has an introductory sentence that clearly describes the topic.

Does not include all of the major sections.

Does not include information not included in the article.

Is concise.

Content:

Content is relevant to the topic.

Content is up to date.

There may be content missing.

Does not deal with Wikipedia equality gaps.

Tone and Balance:

The article is from a neutral point of view.

There are no claims that appear to be heavily biased.

There are some sections which seem underrepresented.

The article does not attempt persuade the reader in favor to one position or away from another.

Sources and references:

Not all facts are backed by reliable secondary sources.

Not all sources are thorough.

The sources appear to be current.

The sources are from a diverse spectrum of authors.

Summary:

The article being summarized is Artificial Neural Networks. The subject is the processing of information by a computer in the same manner it is processed in the human brain. The topic covers the analogy of the human neural networks and networks modeled in computer software.

The history of the artificial neural network is described. The history of using a model of the human neural network began in the early 1940s. It progressed steadily as increases in computer processing increased.

The article describes the models used in artificial neural networks. There is an introduction to the individual components on an artificial neural network. The combination of the individual components to make a functional unit is also explained.

The purpose of training artificial neural networks is discussed. The various methods of training are described. The process of each of the methods are detailed. The environment each of the different learning methods are most useful is provided. A section on applications provides more detail on use of different learning methods.

The article explains some of the theoretical aspects of artificial neural networks. These include computational power, statistics, and the ability of a model to converge on a solution.

Finally, the criticism of artificial neural networks is presented. These include such things as theory, training, and processing. The criticism of the theory focuses on the fact that the artificial neural network does not work exactly like the biological neural network. The criticism of training is the large amount of data required and time to training the network. The criticism of hardware describes the large amount of memory and processing power required for artificial neural networks.