Nueral Networks V. Expert Systems

A neural network is a system of programs and data structures that approximates the operation of the human brain. A neural network usually involves a large number of processors operating in parallel. Each part has its own small specialty of knowledge and access to data in its local memory. It works the same way as the biological nervous system works. This type of system learns by example. It is configured for a specific application to do things like recognize patterns or classify data. It can be thought of as an expert in the category of information and its much easier answer what if questions. A neural network is trained or fed large amounts of data and rules about data relationships. Neural Networks can extract patterns and detect trends that are too complex for humans to take in. Neural networks are sometimes described in terms of knowledge layers. The more complex networks have deeper layers. In feed forward systems the learned relationships about data can "feed forward" to higher layers of knowledge. Examples of where you might use a neural networking is in sales forecasting, risk management, or target marketing. In the business world neural networks can be used in a variety of ways as well.


An Expert System is a system that attempts to mimic the knowledge and decisions of a human. This type of system needs to be created from the knowledge of a human expert. This is an artificial intelligence based system that converts the knowledge of an expert in a specific subject into a software code. This code can be merged with other such codes based on the knowledge of other experts and used for answering questions or queries submitted through a computer. In today's society expert systems attempt to solve problems. There are three different parts that an expert systems typically consists of. First, a knowledge base which contains the information acquired by interviewing experts, and logic rules that govern how that information is applied. Second, an interference engine that interprets the submitted problem against the rules and logic of information stored in the knowledge base. Third, interface that allows the user to express the problem in a human language such as English. The system must be a specialist, justify its conclusions, be able to learn, and estimate the reliability of its answer.   

The difference between the two is quite simple. Expert systems technology is generally regarded as simpler and more widely used than neural network technology. An expert system uses sets of rules and data to produce a decision or recommendation. Neural networks, on the other hand, attempt to simulate the human brain by collecting and processing data for the purpose of “remembering” or “learning”. The primary difference between an expert system and a neural network is that a neural network can adapt its criteria to better match the data it analyzes, while an expert system produces results without adjusting for changes in the analyzed data. Many resources for both technologies can be found on the Internet, from simple explanations to elaborate demonstrations.

Neural Network

When it comes to information technology, a neural network is a system of programs and data structures that approximates the operation of the human brain. A neural network usually involves a large number of processors operating in parallel, each with its own small sphere of knowledge and access to data in its local memory. It works the same way as the biological nervous system works. This type of system learns by example. It is configured for a specific application to do things like recognize patterns of classify data. It can be thought of as an expert in the category of information that it is given, and much easier answer what if questions.

Typically, a neural network is initially trained or fed large amounts of data and rules about data relationships. Neural Networks can extract patterns and detect trends that are too complex for humans to notice. For example, "A grandfather is older than a person's father". A program can then tell the network how to behave in response to an external stimulus's. For example, to input from a computer user who is interacting with the network, or can initiate activity on its own within the limits of its access to the external world. 

In making determinations, neural networks use several principles including gradient-based training, fuzzy logic, genetic algorithms, and Bayesian methods. Neural networks are sometimes described in terms of knowledge layers. With more complex networks having deeper layers. In feed forward systems, learned relationships about data can "feed forward" to higher layers of knowledge. Neural networks can also learn temporal concepts and have been widely used in signal processing and time series analysis.

There are advantages and disadvantages to using neural networks. There is no "perfect" machine learning method. For every problem, for which a certain method is good, there is another problem for which the same method will fail. The methods at which it fails may be solved by other methods quite easily however. Neural networks are quite simple to implement. Although, neural networks cannot be retrained. If you add data later, this is almost impossible to add to an existing network. Also, neural networks often exhibit patterns similar to those exhibited by humans. However this is more of interest in cognitive sciences than for practical examples.

Current applications of neural networks include: oil exploration data analysis, weather prediction, the interpretation of nucleotide sequences in biology labs, and the exploration of models of thinking and consciousness. What makes neural networks more valuable than traditional computing process is their ability to learn. Neural networks offer a different way to analyze data, and to recognize patterns within that data, than traditional computing methods. However, they are not a solution for all computing problems. Traditional computing methods work well for problems that can be well characterized. Balancing checkbooks, keeping ledgers, and keeping tabs of inventory are well defined and do not require the special characteristics of neural networks.

Minskys Lecture

Marvin Minsky is a professor who lectures and teaches about Media Arts and Sciences. He is a professor of Electrical Engineering and Computer Science Description. In the lecture i listened too, Marvin talked about common sense thinking, artificial intelligence, and the future of the human mind. There is a long history of artificial intelligence. Minsky discussed the history, the future of the growth of artificial intelligence as well as the current situation, and where it will find its place in the coming decades. Human beings have a logical way of thinking, and Minsky discussed why it is hard to create machines aka computers, databases, etc, using artificial intelligence to match the human mind.

Minsky asked the question, Does an expert system contain common sense? In his opinion, machines may have a little common sense. Machines store a certain knowledge base which in return allows the common sense. There is certain things that are able to assist the machines to have some answers, but not all. Just because they have a knowledge base, does not make them experts. Machines can be created for specific purposes or tasks but it is the creation that allows this, not just expert knowledge. This is different than huamns because we do have common sense even if we have no expertise. Minsky used a specific example to illustrate this idea. He used a 4 year old child being able to identify things in a room. This is common sense for the kid, and it does not require the kid to operate off of different codes like machines do. If you asked a machine to identify objects in a room, it might be impossible because they need to operate off of codes. They need specific instructions, but a child can learn from the environment and do so quite easily and fast. It takes a lot of time and intelligence to program these machines to be able to perform actions, because they can not think on their own with common sense. In the future we may be able to make a machine which inherits lots of common sense to be able to do something such as sweing up a torn shirt. At the moment, there is nothing like this in the world.