|
FROM: Wai Shing Chan
DATE: May 15, 2000 TIME: 15:50:17 REMOTE IP: 163.238.34.212 |
Artificial Intelligence |
Artificial Intelligence applies into financial institutions Today, artificial intelligence technology maybe applied for positive identifications of individuals during financial transactions, such as automated banking transactions, telephone transactions etc. This study focuses on academic research in neural network technology. What is a neural network? In the AI community, the proponents of expert systems have approached the challenge of simulating intelligence differently that their counterpart proponents of neural networks. Expert systems contain the coded knowledge of a human expert in a field; this knowledge takes the form of “if-then” rules. The problem with this approach is that people do not always know why they do what they do. And even when they can express this knowledge, it is not easy to translate into usable computer code. Also expert systems are usually bound by a rigid set of inflexible rules which do not change with experience gained by trail and error. In contrasts, neural networks are designed around the structure of a biological model of the brain. Neural networks are composed of simple components called “neurons” each having simple tasks, and simultaneously communicating with each other by complex interconnections. As Herb Brody states, “Neural networks do not require an explicit set of rules. The network-rather like a child-makes up its own rules that match the data it receives to the result it’s told is correct”(Computers that Learn by Doing, P.42). Ability to learn by example is the characteristic of neural networks that makes them best suited to simulate human behavior. Computer scientists have exploited this system characteristic to achieve breakthroughs in computer vision, speech recognition, and optical character recognition. Neural networks restructure their knowledge base at each step in the learning process. Much of the technology is currently in the research phase and has yet to produce a commercially available product, such as visual recognition applications. History of Artificial Intelligence… the study of the brain was once limited to the study of living tissue. Any attempts at an electronic simulation were brushed aside by the neurobiologist community as abstract conceptions that bore little relationship to reality. This was partially due to the over-excitement in the 1950’s and 1960’s for networks that could recognize some patterns, but were limited in their learning abilities because of hardware limitations. In the 1990’s computer simulations of brain functions are gaining respect as the simulations increase their abilities to predict the behavior of the nervous system. This respect is illustrated by the fact that many neurobiologists are increasingly moving toward neural network type simulations. Search result: Sejnowski, one of neurobiologist has introduced a three-layer net, which has made some excellent predictions about how biological systems behave. This network consisting of three layers, in which a middle layer of units connects the input and output layers. When the network is given input, it sends signals through the middle layer which checks for correct output. An algorithm used in the middle layer reduces errors by strengthening or weakening connections in the network. This system, in which the system learns to adapt to the changing conditions, is called back-propagation. The Sejnowski’s network has tested by a team at the Massachusetts Institute of Technology and the result was nearly identical to what they found in their experiments with monkeys. Types of Synthesized Senses- 1) Visual Recognition… This ability of a computer to distinguish one customer from another is not yet a reality. But, recent breakthroughs in neural network visual technology are bringing us closer to the time when computers will positively identify a person. Studying the retina of the eyes is the focus of research by two professors at the California Institute of Technology, Misha A. Mahowlad and Carver Mead. Their objective is to electronically mimic the function of the retina of the human eye. Previous research in this field consisted of processing the absolute value of the illumination at each point on an object, and requires a very powerful computer (Overturning the Category Buscket, P.249-250). The analysis required measurements be taken over a massive number of sample locations on the object, and so, it requires the computing power of a massive digital computer to analyze the data. They believe that to replicate the function of the human retina they can use a neural network modeled with a similar biological structure of eye rather than using massive computer power. Their chip utilizes an analog computer, which is less powerful that the pervious digital computes. They compensated for the reduced computing power by employing a far more sophisticated neural network to interpret the signals from the electronic eye. They modeled the network in their silicon chip based one the top three layers of the retina which are the best understood portions of the eye. There are the photoreceptors, horizontal cells, and bipolar cells. The electronic photoreceptors (first layer) will accept incoming light and transform it into electrical signals. Then, the horizontal cell (second layer) uses a neural network technique by interconnecting second layer and the bipolar cells (third layer). The connected cells then evaluate the estimated reliability of the other cells and give a weighted average o the potentials of the cells around it. Nearby cells are given the most weight and far cells less weight. This technique is very important to this process because of the dynamic nature of image processing. If the image is accepted without testing its probable accuracy, the likelihood of image distortion would increase as the image changed. They developed a silicon chip that contains about 2,500 pixels—photoreceptors and their associated image-processing circuitry. The accuracy of this silicon chip displays the usefulness of analog computing despite the assumption that only digital computing can provide the accuracy necessary for the processing of information. As close as these systems come to imitating their biological counterparts, they still have a long way to go. For a computer to identify more complex shapes, the professors estimate the requirements would be at least 100 times more pixels as well as additional circuits that mimic the movement-sensitive and edge-enhancing functions of eye. 2) Voice recognition.. it is another area that has been the subject of neural network research. Researchers have long been interested in developing an accurate computer-based system capable of understanding human speech as well as accurately identifying one speaker from another. Ben Yuhas, a computer engineer at John Hopkins University, has developed a promising system for understanding speech and identifying voices that utilizes the power of neural networks. Previous attempts at this task have yielded systems that are capable of recognizing up to 10,000 words, but only when each word is spoken slowly in an otherwise silent setting. Unfortunately, this type of system is easily confused by back ground noise. Yuhas' theory is based on the notion that understanding human speech is aided, to some small degree, by reading lips while trying to listen. The emphasis on lip reading is thought to increase as the surrounding noise levels increase. This theory has been applied to speech recognition by adding a system that allows the computer to view the speaker’s lips through a video analysis system while hearing the speech. The computer, through the neural network, can learn from its mistakes through a training session. Looking at silent video stills of people saying each individual vowel, the network developed a series of images of the different mouth, lip, teeth, and tongue positions. It then compared the video images with the possible sound frequencies and guessed which combination was best. Yuhas then combined the video recognition with the speech recognition systems and input a video frame along with speech that had background noise. The system then estimated the possible sound frequencies from the video and combined the estimates with the actual sound signals. After about 500 trial runs the system was as proficient as a human looking at the same video sequences. This combination of speech recognition and video imaging substantially increases the security factor by not only recognizing a large vocabulary, but also by identifying the individual customer using the system. This system can use for voice-activated dialing for cellular phones, made secure by their recognition and authorization of a single approved caller. International telephone companies such as Sprint are using similar voice recognition systems. Conclusion… Neural networks are still considered emerging technology and have a long way to go toward achieving their goals. This is certainly true for financial transaction security. But with the current capabilities, neural networks can certainly assist humans in complex tasks where large amounts of data need to be analyzed. For visual recognition, neural networks are still in the simple pattern matching stages and will need more development before commercially acceptable products are available. Speech recognition, on the other hand, is already a huge industry with customer ranging from individual computer users to international telephone companies. For security, voice recognition could be an added link to the chain of pre-established systems. For example, automated account inquiry, by telephone is a popular method for customers to determine the status of existing accounts. With voice identification of customers, an option could be added for a customer to request account transactions and payments to other institutions. For credit card fraud detection, banks have relied on computers to identify suspicious transactions. I was looking forward that all crimes of financial transaction are eliminating by applying the neural network technology.
It must be fully test or bug free. It is because if some major industry use it and it crash, it will become a large problem. It might be relate to another industry too.
REPLY TO THIS ARTICLE
Email Professor Klibaner with any class questions
Email Han with any questions/problems regarding this site
Re: Artificial Intelligence
Posted by Bougataya Mehdi on May 30, 2000 at 12:41:45 at IP 172.147.168.64
Actually I think that AI is helping as in alot of different field, as voyager when was sent in the outer space it was self manuvered, because if it wasn't the time it takes to send a message from saturn and response to it is about 50 min. So it is hard to control something with that much time difference. So due to AI the satellite was manuvering it self. The same story to the satelite sent to Mars Re: Artificial Intelligence
Posted by Bobbie Razo on May 28, 2000 at 18:28:49 at IP 216.192.103.36
Neural network is a very promising field in AI technology. Since this field is progressing to develop human-like applications and robots, we also have to be cautious in the future that it doesn’t become a tool and target for hostile purposes like the viruses and the internet. Re: Artificial Intelligence
Posted by Joseph Fardella on May 23, 2000 at 23:46:25 at IP 63.23.128.165
I do not believe that AI will ever control humans. The creater never implements a system that he/she cannot pull the plug on. It is because of this that there will always be safegaurds to protect from such a situation. Re: Artificial Intelligence
Posted by Kam Tsui on May 23, 2000 at 23:09:41 at IP 4.54.120.229
If people program AI morally, it would be convininent for human beings; otherwise, it'll end up AI controlling us like in the fiction. Re: Artificial Intelligence
Posted by Hector Blanco on May 23, 2000 at 1:39:54 at IP 172.135.166.34
It will be very interesting to see how far technology can be pushed with the use of artificial intelligence. Re: Artificial Intelligence
Posted by Dennis Healy on May 23, 2000 at 1:26:03 at IP 205.188.199.32
Artificial intelligence is probably going to be the biggest field in the future. One of the biggest problems with the internet is that it is too difficult to use. This may sound funny to you as a CS major, but there are many people out there who have very little or no computer experience(many older people complain that Bank ATMs are too dificult to use). As the computers and programs get smarter, they will be easier to use and reach more people. Re: Artificial Intelligence
Posted by mkhanum@hotmail.com on May 22, 2000 at 17:15:50 at IP 163.238.9.13
Artificial Intelligence really amazes me. It is so great to have a machine do what we wish for it to do. It still has a long way to go. It is close to impossible to follow human neural networks in human brain and implement it in a machine. However, the course that CSI offer has created some interesting robusts. I wish them good luck. Re: Artificial Intelligence
Posted by Jick Gee Chui on May 18, 2000 at 17:26:35 at IP 163.238.34.214
It still is a new technology. Re: Artificial Intelligence
Posted by Han on May 18, 2000 at 15:13:44 at IP 198.83.28.35
There is an online book about this subject, located here: Re: Artificial Intelligence
Posted by Han on May 18, 2000 at 14:50:08 at IP 198.83.28.35
Could the Internet at large be set up as a huge neural net? Re: Artificial Intelligence
Posted by Kit Woo on May 18, 2000 at 0:48:13 at IP 152.172.131.138
I think that neural networks have a long way to go. The idea for the implementation in the financial industry is great, but I think the majority of people is very conservative and does not want to risk their money. They do not trust the new technology when their moneies are involve. Just go to any bank and see how many people carry their passbook savings.
