G. F. Luger and W. A. Stubblefield, Artificial Intelligence and the Design of Expert Systems. Redwood City, CA: Benjamin/Cummings, 1989.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....axis. This metric is easy to use and has been shown to work well in our experiments presented in Section 7. 5.4 SEARCHING FOR A SHORTEST PATH The second step in the algorithm is to find a shortest path in 6 be tween qinit and qgoal, or determine that none exists. We use the A algorithm [33], and a metric Ppath : x [0, cx ) to measure the length of a path and the remaining distance to qgoal If the search procedure succeeds in finding a path, we need to check it for collision. Otherwise, if no path exists in the roadmap, we either report failure, or go to the node enhancement step ....

G.F. Luger and W.A. Stubblefield. Artificial intelligence and the design of expert systems. Benjamin/Cummings, Redwood City, CA, 1989.

....computer system configuration, action planning and game playing are typical KBS application domains. A number of programming paradigms have been developed which support the construction of KBS software, one of the most widespread of which is production system technology, e.g. Luger Stubblefield [Luger Stubblefield, 1989]. Typically, a production system involves a collection of data items, referred to as working memory, a collection of rules, a control strategy and a conflict resolution strategy. The control strategy defines a mechanism by which a number of rules are selected in order to modify or deduce ....

Luger G. F. & Stubblefield W. A. 1989 Artificial Intelligence and the Design of Expert Systems. The Benjamin Cummings Series in Artificial Intelligence.

....a modification of non deterministic conditional ground term rewriting systems (Jounanaud Dershowitz, 1990) where we extend terms with some useful features for pattern matching and extend the matching process using equational theories. See (Beynon Davies, 1993) Lucas 4 Van Der Gaag, 1991) or (Luger Stubblefield, 1989) for a general introduction to rule based systems; see (Goguen, 1976) for an introduction to equational algebra. 2.1 Syntax and semantics Let V and F be two disjoint sets of variables and function symbols respectively. Each function symbol f 2 F has a positive integer arity arity(f ) ....

Luger G. F. & Stubblefield W. A. (1989) Artificial Intelligence and the Design of Expert Systems. The Benjamin Cummings Series in Artificial Intelligence.

....the resource becoming otherwise available) The basic matcher is a very simple resource allocation control program, and we can easily imagine using other more sophisticated algorithms for special applications. This same mechanism will clearly apply to blackboard systems in Artificial Intelligence [58], since they are also global data spaces with multiple access from parallel sources. Actually, any collection of interacting processes that communicate via a language with fixed syntax can be interpreted this way: the advantage of the interpretation is increased flexibility in the responses to the ....

George F. Luger, William A. Stubblefield, Artificial Intelligence and the Design of Expert Systems, Benjamin / Cummings (1989)

....Rule based systems (often because of smaller scope) tend to have deeper nestings of rules, with conclusions extending many rules out from the source. In cases where there is some uncertainty or error in the rules, this results in a quickly decreasing reliability as the chain extends(see [30], especially the section on MYCIN) Because of its tradeoff towards a broader selection of experts, ZWrap has yet to encounter issues inherent in extensive rule chaining. Lastly, there is the issue of topology (see Figure 2 5) With a rule based system, any of several rules may result in the ....

George F. Luger and William A. Stubblefield. Artificial Intelligence and the Design of Expert Systems. The Benjamin/Cummings Publishing Company, Inc., 1989.

....input is a short segment of DNA nucleotides, and the goal is to learn to predict whether the DNA segments contain a promoter site. There are 31 rules in the domain theory, and 940 examples in the dataset. financial advisor The financial advisor rule base contains 9 rules as shown in Figure 2 [19]. As in [5] a set of 5500 labeled examples that are consistent with the rule base is randomly generated. 500 examples are used for training and the remaining 5000 is used for testing. 1 if (sav adeq and inc adeq) then invest stocks 2 if dep sav adeq then sav adeq 3 if assets hi then sav adeq ....

Luger, G. F., and Stubblefield, W. A. 1989. Artificial Intelligence and the Design of Expert Systems. Redwood City, CA: Benjamin/Cummings.

....management. 2. User Centered Model of Hypertext Interaction We propose to support the usercentered hypertext navigation with techniques drawn from the areas of natural language processing (Gazdar Mellish 1989) computer aided learning (Burns Capps, 1988) and knowledge based systems (Luger Stubblefield, 1989). Our method relies on a view of hypertext navigation to be a tutorial dialog between two intelligent agents, the user being a student and the hypertext system being a tutor. The main objective of the student is to acquire maximum information about a certain topic of discourse, while the tutor ....

Luger, G. F., & Stubblefield, W. A. (1989). Artificial intelligence and the design of expert systems. Redwood City, CA: Benjamin/Cummings.

....are also made to work in other related fields such as that of knowledge acquisition. 2 Review of Relevant AI Techniques 2. 1 Introduction Artificial intelligence (AI) has been defined as: the branch of computer science that is concerned with the automation of intelligent behaviour [LS89] A discussion of the exact definition of intelligent behaviour is beyond the scope of this report. Some feeling for what is intended may, however, be obtained from a brief characterisation of the field of AI in terms of the kinds of approaches most often used and the problems which are ....

....that it is modular and therefore flexible in that parts of it can easily be changed or updated. A number of references to systems using rule based representations of knowledge will be made in the following sections. Further discussion on the subject of rules can be found in [Ric83, Mis88] and [LS89] We now move on to consider frame based representations. 2.2.4 Frames This section describes a representation scheme based on the use of frames or schemata, with examples derived from those in [LS89] This representation has been attacked as adding nothing to the tools of AI, but remains ....

[Article contains additional citation context not shown here]

G. F. Luger and W. A. Stubblefield. Artificial Intelligence and the Design of Expert Systems. Benjamin Cummings, 1989. 55

....only by pre existing set of rules. In our hybrid system the domain theory is transformed into an initial network through an extended version of KBANN s symbolic knowledge encoding. As an example, consider the rule base in Table 1 which is a modified version of the simple financial advisor from [10]. if (savings adequate and income adequate) then invest stocks if dependent savings adequate then savings adequate if assets high then savings adequate if (dependent income adequate and earnings steady) then income adequate if debt low then income adequate if (savings dependents Theta 5000) then ....

G. F. Luger and W. A. Stubblefield. Artificial Intelligence and the Design of Expert Systems. Benjamin/Cummings, Redwood City, CA, 1989.

....Financial Advising Model The multistrategy classification systems are tested on a small financial advising problem in which the objective is to advise users whether to invest in the stock market or not, based on several financial parameters. The problem is represented by a rule base adapted from [7] and shown in Table 1. Although this model is extremely simplified, it illustrates issues involved in realistic financial advising. The system s input consists of six real variables, shown in italics in Table 1. The system output invest stocks is a binary variable, representing the recommendation ....

G.F. Luger and W.A. Stubblefield, Artificial Intelligence and the Design of Expert Systems, Benjamin /Cummings, 1989.

....that two approaches to classification which do not perform well independently may result in an improved classification integrated into in a hybrid system. Figure 4.6: Two Spirals: All Hyperplanes 4.6. Financial advisor This problem is a modified version of the simple financial advisor from [30]. The task is to determine whether an individual is to invest capital in additional savings or in the stock market. Six features are input: annual income, if the income source is steady, current assets, current savings, annual debt payments and the number of dependents. The expert system rule base ....

G. F. Luger and W. A. Stubblefield. Artificial Intelligence and the Design of Expert Systems. Benjamin/Cummings, Redwood City, CA, 1989.

....in the Neural Network The symbolic knowledge encoding procedure translates a domain theory in the form of propositional rules into a network of TLU [4] 12] The weights of the individual TLU are chosen to satisfy the rules. For example, consider the simple financial advisor rule base (due to [16]) in Table I. if (sav adeq and inc adeq) then invest stocks if dep sav adeq then sav adeq if assets hi then sav adeq if (dep inc adeq and earn steady) then inc adeq if debt lo then inc adeq if (sav dep 5000) then dep sav adeq if (assets income 10) then assets hi if (income 25000 dep ....

G. F. Luger and W. A. Stubblefield, Artificial Intelligence and the Design of Expert Systems, Benjamin/Cummings, Redwood City, CA, 1989.

....data is used again to train the perceptron. The random partitions were repeated 100 times for each of the 3 real world data sets and the overall misclassification rates were calculated. The four data sets are described as follows: Advisor The data are generated from a rule base adapted from (Luger Stubblefield 1989) and is shown in Table 2. It is a simplified system trying to illustrate issues involved in real life financial advising. The system consists of 5 input variables, shown in bold in Table 2, and a binary response variable invest stocks. We simulated 1,000 patterns for the classifiers training set, ....

Luger, G. F. and Stubblefield, W. A. 1989. Artificial Intelligence and the Design of Expert Systems.

....USA and Department of Computer Science, University of New Mexico, Albuquerque NM 87131, USA. Email: terry santafe.edu 1 Introduction The field of Artificial Intelligence (AI) has long recognized the importance of structure in search. Most of the classic AI search algorithms, see, for example, [6, 21, 30, 40], are either explicitly phrased as algorithms that search graphs, or can be described in terms of a search on a graph. The other component of search is navigation: the process of examining a structure for some purpose. Myriad navigation strategies, from random search to exhaustive search, have ....

Luger, George F. and Stubblefield, A. [1989] Artificial intelligence and the design of expert systems, Benjamin/Cummins, Redwood City CA.

....depth here because of the overwhelming amount of details and variations. They are analyzed in [6] To perform approximate matching one must have a model to store the information. Knowledge representation has been studied for years but there is no general consensus or solution to the problem [2, 7, 10, 11, 14]. It is generally believed that a good knowledge representation model should possess the quality of generality to be able to represent very diversified real world situations, extensibility to accommodate the varying amount of information from one case to another, and decomposability to allow the ....

Luger G. and Stubblefield W.: `Artificial Intelligence and the Design of Expert Systems', Benjamin/Cummings, 1989.

No context found.

G. F. Luger and W. A. Stubblefield, Artificial Intelligence and the Design of Expert Systems. Redwood City, CA: Benjamin/Cummings, 1989.

No context found.

Luger, G. F. and Stubblefield, W. A. (1989), Artificial Intelligence and the Design of Expert Systems, Benjamin/Cummings Publishing (ISBN 080530139 -9).

No context found.

Luger GF, Stubblefield WA. Artificial intelligence and the design of expert systems. Benjamin Cummings, Redwood City, CA, 1989

No context found.

G. Luger and W. Stubblefield, "Artificial Intelligence and the Design of Expert Systems, " The Benjamin/Cummings Publishing Company, Inc. 1989.

No context found.

Luger, G.F. and W.A. Stubblefield 1989. Artificial Intelligence and the Design of Expert Systems, Redwood City, CA:The Benjamin/Cummings Publishing Company, Inc.

No context found.

Luger, George F. & Stubblefield, William A., "Artificial Intelligence and the Design of Expert Systems", The Benjamin/Cummings Publishing Company, Inc., 1989

No context found.

Luger, G. F. and W. A. Stubblefield. 1989. Artificial intelligence and the design of expert systems, Benjamin/Cummings Pub. Co., Redwood City, CA.

No context found.

Luger, G. F.; Stubblefield, W. A. Artificial Intelligence and the Design of Expert Systems. Redwood City, CA: Benjamin/Cummins; 1989.

No context found.

Luger, George F., and William A. Stubblefield. Artificial Intelligence and the Design of Expert Systems . Benjamin Cummings, 1989.

No context found.

G. F. Luger and A. Stubblefield. Artificial Intelligence and the Design of Expert Systems. Benjamin/Cummins, Redwood City CA, 2nd edition, 1993.

First 50 documents

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC