Uzman sistem teknolojisi ve çizelgeleme uygulaması
The technology of expert system and applications of scheduling
- Tez No: 19355
- Danışmanlar: Y.DOÇ.DR. ORHAN KURUÜZÜM
- Tez Türü: Yüksek Lisans
- Konular: Bilgisayar Mühendisliği Bilimleri-Bilgisayar ve Kontrol, Computer Engineering and Computer Science and Control
- Anahtar Kelimeler: Belirtilmemiş.
- Yıl: 1991
- Dil: Türkçe
- Üniversite: İstanbul Teknik Üniversitesi
- Enstitü: Fen Bilimleri Enstitüsü
- Ana Bilim Dalı: Belirtilmemiş.
- Bilim Dalı: Belirtilmemiş.
- Sayfa Sayısı: 155
Özet
Yapay zeka bilgisayar yardimıyla zekice davranışların. incelenme yollarından birisidir. Diğer anlamda bilgisayara zekice düşündürme yolarını bulmaya çalışır. însan zekasının sınırlı Ölçüde bilgisayara aktarımıdır. Uzman sistemler bir çeşit yapay zeka programıdırlar. Amacı insan tecrübe ve bilgilerini otomasyona tabi tutup, kaybolmalarını önlemektir. Fakat bir insanın tecrübelerini makinaya aktarmak oldukça zordur, çünkü tecrübelerin hepsini özellikle sezgiye dayanan tecrübeleri makinaya aktarmak büyük zorluklar gerektirir. Fakat optimal aktarımı sağlayacak bazı teknikler bilgi mühendisi tarafından ortaya konulmuştur. Bu teknikler bilgi mühendisleri tarafında oldukça etkin bir şekilde kul 1 anılır, Bilgi mühendisi uzman ve makina arasında daha doğrusu programcı arasında köprü vazifesi görmektedir. Görevi, uzmandan bilgiyi eni yi şekilde almaya çalışmak ve alınan bilgiyi belli standartlar halinde programcıya sunmaktır. Bilgiyi belli kalıplara sokmak, problem çözme stratejileri geliştirmek bilgi mühendisinin başlıca görevleri arasındadır. Yukarıda da belirtildiği gibi bir uzman sistemi inşa etmek için üç önemli elemana ihtiyaç vardır. Birincisi bir veya daha fazla uzman, ikincisi bilgi mühendisi, üçüncüsü ise programcıdır. Uzman sayısının fazla olması, uzmanların farklı tecrübelere sahip olmaları yada aynı doğruya farklı yollardan ulaşmaları açısından birtakım sorunlar ortaya çıkaracaktır. Fakat iyi bir bilgi mühendisi uzmanlardan faydalanabildiği ölçüde uzman sistemi geliştirecektir. Uzman sistemler belirli aşamalardan geçilerek inşa edilir. Uzman sistem inşa etmek için konvansiyonel programlama dillerinden faydalanılabilir. Fakat mevcut yapay zeka dilleriyle uzman sistemi inşa etmek daha hızlı ve daha kolay olacaktır. Burada LİSP yada PROLOG gibi yapay zeka dillerini iyi bilen programcı bulmanın zorluğunu ve maliyetini de belirtmek gerekir. Fakat uzman sistemi daha hızlı ve daha az maliyetli inşa etmenin en iyi yolu uzman sistem shell' leridir. Shell 'ler çok kolay uzman sistem inşa etmemeizi sağlarlar fakat esneklik açısından bakıldığında LÎSP ve PROLOG dillerinin yerini tutamazlar. Spesifik konular için birçok. uzman sistem shell 'i inşa edilmiştir. Tıp, Mühendislik, mimarlık gibi alanlarda oldukça fazla shell bulunmaktadır. Bu alanlarda her konu için bir shell bulunduğunu söylesek herhalde abartmış olmayız. Özetle söylemek gerekirse uzman sistemler geleceğin sistemleridir, gelecek yıllarda 'her alana girmeleri büyük olasılıkla mümkündür.
Özet (Çeviri)
Artificial Intelligence is a simply a way of making a computer think mtel I igent iy. This is accomplished by studying how people tnink when they are trying to make decision and solve oroülams, breaking those thought processes down into Dasic steps, and designing a computer pragram.that solves problems using those same steps, Artificial Intelligence thereby provides a simple, structured approach to designing complex decision-making programs. Standart computer program can only provide answers to problems for which it is specifically programmed. If standart program needs to be modified in order to accommodate new information, the entire program may have to be scanned until the optimum space is found to insert the modification. This is not only time consuming» but other parts of the program may be adversely affected in the process and errors may result. Artificial intelligence, as its name implies, really does enable a computer to think. By simplifying 'the way programs are put together. Artificial intelligence imitates the oasic human learning process by which new information is absorbed and made available for future reference. The human mind can incorporate new knowledge without Changing the way the mind works or disturbing all the other facts that are already stored in the brain. An Artificial intelligence program works in very much the same way. Artificial intelligence techniques allow the construction of a program represent a bigly independent and identifiable step toward tne solution of a. problem of set of problems. Let ? s consider this carefully. Each oiece of the program is like a piece of information in a persons mind. If that information is disputed» the mind can' automat icai ly adjust its thinking to accommodate a new set of facts. One doesn't have to go ebaut reconsidering every piece of information one nas ever learned, only those few pieces that are relevant to particular change, A standart program can do every think an artificial intelligence program can do, but it cannot be programmed as easily or as quicly. En both types of programs, all nieces are i nter dependent in tne way they carry out their designed function. But: an artificial intelligence program possesses a notable characteristic wnich is equivalent to a vital characteristic of human intelligence. Each minute piece can be modified withoutaf f ecting the structure of the entire program. This flexibility provides greater programming efficiency and understandability - in a word, intelligence. Since artificial intelligence is a science rooted in human thought processes, an examination of how people think essential. Of course, no one knows exactly how the mind works. Human intelligence is a complex function that scientists have only begun to understand, but enough is known for. us to make certain assumptions ebaut how we think and apply those assumption in designing artificial intelligence programs. All thinking helps us accomplish something. When the alarm clock rings in the morning, a thought process must be employed to guide your hand to the button to turn it off. It isn't an automatic reaction ; a specif ik response was sought to solve a particular problem. The final results to which all our thought processes are directed are called goals. Once you have reached the goal of turning off the alarm, your mind is immediately confronted whit other goals to be reached, such as getting to the bathroom, brushing your teeth, getting dressed, making and eating breakfast, going to bus stop, and so on. These are all goals which, when accomplished, lead you to the ultimate goal of getting to work and getting there on time. None of the thoughts guiding you to this final results are random or arbitrary. They have been pressed into service because every step of the way you had a specific goal in mind. When engaged in the most simple physical task or the most complex mental activity, the mind is sharply-focused on a goal. Without goals, we have no reason to think. We all know that the human mind possesses a vast store of knowledge relating to a countless array of objects and ideas. Survival depends on our ability to apply this knowledge to any situation that arises and to continuously learn from new experiences so that we will be able to respond to similar situations in the future. What is generally considered to be intelligence can be broken down into a collection of facts and means of utulising these facts to reach goals. This is done, in part, by formulating sets of rules relating to all the facts stored in the brain.Expert systems are kind of artificial intelligence program. Expert systems are computer programs that are constructed to do kinds of activities that human experts can do, such as design, compose, plan, diagnose, interpret, summarize, audit, give advice. The work such a system is concerned with is typically a task from the worlds of business or engineering/science or government. Expert system programs are usually set up to operate in a manner that will be perceived as intelligence; that is, as if there were a human expert on the other side of the video terminal. A characteristic body of programming techniques give these programs thair power. Expert systems generally use automated reasoning and the so-called weak methods, such as search or heuristic, to do their work. These techniques are quite distinc from the well-articulated algorithms and crisp mathematical procedures of more traditional programming. Expert systems occupy a narrow but very important corner of the entire programming establishment. As part of saying what they are, we need to describe their place within the surrounding framework of established programming systems. Expert systems are part of a larger unit we might callartif icial programming. Procedural programming is what everyone learns when they first begin to use BASIC or PASCAL or FORTRAN. Procedural programming and artificial intelligence programming are quite different in what they try to do and in how they try to do it. In traditional programming (procedurel programming), the computer has to be told in great detail exactly what to do and how to do it. This style has been very successful for problems that are well defined.. They usually are found in data processing or in engineering or scientific work. Artificial intelligence programming sometimes seems to have been defined by default, as anything that goes beyond what is easy to do in traditional procedural programs, but there are common elements in most artificial intelligence program. What characterizes these kinds of programs is that they deal with complex problems that are often poorly understood, for which there is nocrisp algoritmic solution, and that can benefit from some sort of symbolic reasoning. There are substantial differences in internal mechanisms of the computer languages used for these sorts of problems. Prosedural programming focuses on the use of the assignment statement for moving data to and from fixed, prearrenged, named locations in memory. These named locations are the program variables. It also depends on a characteristic group of control constructs that tell the computer what to do. Artificial intelligence programs are usually written in languages like LISP AND PROLOG. Program variables in these language have an ephemeral existence on the stack of the underlying computer rather than in fixed memory locations. Data manipulation is done through pattern matching and list building. The list techniques are deceptively simple, but almost any data structure can be built upon this foundation. Many examples of list building will be seen later when we begin tu use Prolog. Artificial intelligence programs also use different set of control constructs. Expert systems are really distinguished from other artificial intelligence by their aim and by their packaging. By aim we mean that the expert system attempts to do an activity that is appropriate for a human expert. By packaging we mean there is usually an attempt to make the program seem intelligence rather then merely mechanical.Particulars that enhance this appearance are if the inner functioning of the program mirrors the way a human would attack the problem, if the problem can explain its action in a way that would make sense to a human being, and if it can interact with the operator by using an unstilted, flexible dialog that is similar to natural speech.- Expert system use all of the programming techniques that appear in other artificial intelligence programs, but certain ones tend to be accentuated. In common with other computer projects, well planned and well executed development stages will' lead to expert systems that perform to the standarts required of them. Unlike traditional data processing and information systems though, there has been little time to evolve the clearly documented and methodologies that "characterize the development of traditional systems projects. There is general, though by no means universal, agreement that prototyping offers a useful approach.One of the key factors leading to a succesful expert system is the choise of problem domain within which to work. An area which is rule based and clearly defined, together with the presence of experts willing to aid in the development process, offers the opportunity to benefit from the production of an expert system. These benefits depend on the type of system but might include cost saving for consultation, improvements in and standardization of specialists' performances, the archiving of rare skills, development of a consistent and effective training tool for employees and the possibility of marginal improvements on high volume or large value transactions. -The building of the prototype is facilitated by the choice an appropriate expert system tool, shell or possible artificial intelligence language. Knowledge is elicited by the engineer using a combination techniques. Examples are structured interviewing, protokol analysis, goal decomposition and multi-dimensional scaling. In some cases automated induction on a set of examples may generate appropriate rules. The evaluation of the prototype provides a guide to the knowledge engineering of the success of the represantation and inference strategies chosen. Development of the main system may involve expanding the prototype in both breadth and depth, though often the prototype is forgotten and the final system is built a new using the lessons learnt in prototyping. As well as concentrating on the problem area and the relevant representation, the knowledge engineering needs a model of the level and type of knowledge assumed of the target user in order to define the scope and content of the knowledge base and the system's user interface. Fine tuning of the developed system is necessary at the final stages of the development.. The evaluation of an expert system leads to problems not encountered in the assesment of traditional information system. This results from the large number of logical paths generated by the total space of possible cases the which the system may be applied.Further exacerbation is caused by the presence of uncertainty and the recognation that the expert system need not always be right. Over the next few years it is likely that expert system methodologies will be developed as artificial intelligence in busines becomes increasingly important. I think that expert systems will be used for production planning and control, especially for routing and scheduling.
Benzer Tezler
- An Expert system design for a transportation problem
Başlık çevirisi yok
BURAK ACAR
Yüksek Lisans
İngilizce
1994
Elektrik ve Elektronik MühendisliğiMarmara ÜniversitesiEndüstri Mühendisliği Ana Bilim Dalı
DOÇ. DR. LALE DURUİZ
- Öngörü işlevli kontrol ve Matlab'de uygulama tasarımı
Predictive functional control and application design in Matlab
NESLİHAN KARAŞ
Yüksek Lisans
Türkçe
2019
Bilgisayar Mühendisliği Bilimleri-Bilgisayar ve Kontrolİstanbul Teknik ÜniversitesiKontrol ve Otomasyon Mühendisliği Ana Bilim Dalı
PROF. DR. AFİFE LEYLA GÖREN
- Öğretmenlik uygulaması modeli
Practice teaching model
FATMA HACIOĞLU
Doktora
Türkçe
1990
Eğitim ve ÖğretimAnkara ÜniversitesiEğitim Programları ve Öğretimi Ana Bilim Dalı
PROF. DR. CEVAT ALKAN
- Deniz taşımacılığı için sürdürülebilir biyoyakıt değerlendirme modeli
Sustainable biofuel assessment model for maritime transportation
ÇAĞATAYHAN SEVİM
Yüksek Lisans
Türkçe
2023
Denizcilikİstanbul Teknik ÜniversitesiDeniz Ulaştırma Mühendisliği Ana Bilim Dalı
DR. ÖĞR. ÜYESİ BURAK ZİNCİR
- An Integration of expert system with a mathematical programming model
Bir matematiksel programlama modelinin uzman sistem ile entegrasyonu
ALİ SERDAR TAŞAN
Yüksek Lisans
İngilizce
2000
Endüstri ve Endüstri MühendisliğiDokuz Eylül ÜniversitesiEndüstri Mühendisliği Ana Bilim Dalı
PROF. DR. İREM ÖZKARAHAN