R. S. Michalski and K. A. Kaufman, "Data mining and knowledge discovery: A review of issues and a multistrategy approach," in Machine Learning and Data Mining: Methods and Applications, R. S. Michalski, I. Bratko, and M. Kubat, Eds. New York: Wiley, 1997.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....etc. and suggests modelling decisions. To put our approach into practice the results of the implemented data mining techniques are aligned with the modelling primitives given in our ontology model. This enables the combination of results and realizes a multi strategy learning architecture [20], which supports balancing between the advantages and disadvantages of the different data mining techniques. Our implementation also follows the balanced cooperative modeling paradigm established by Morik [22] Her work describes the interaction between knowledge acquisition and machine learning, ....

R. Michalski and K. Kaufmann. Data mining and knowledge discovery: A review of issues and multistrategy approach. In Machine Learning and Data Mining Methods and Applications. John Wiley, England, 1998.

....of our system. They operate on the extracted information and are used for two tasks: One task is the acquisition of new structures, the second task is the evaluation of given structures. As mentioned before one of the core capabilities of our system is multi strategy learning (as described in [18]) All learning methods use a common result structure. Therefore the engineer can combine results and is supported in balancing between advantages and disadvantages of different learning methods. Due to this combination and balancing the complex task of ontology engineering is fitted better. 1 ....

R. Michalski and K. Kaufmann. Data mining and knowledge discovery: A review of issues and multistrategy approach. In Machine Learning and Data Mining Methods and Applications. John Wiley, England, 1998.

....relations based on frequent couplings of concepts. Combining results is enabled by the implementation of a common result structure for all learning methods. The complex task of ontology engineering is fitted better as it is possible to combine the results from different learning methods. (Michalski and Kaufmann, 1998) describe that multi strategy learning architectures support balancing between advantages and disadvantages of different learning methods. Ontology Engineering In our approachofsemiautomatic ontology acquisition extensive support for ontology engineering is necessary. Manual ontology modeling ....

R. Michalski and K. Kaufmann. 1998. Data mining and knowledge discovery: A review of issues and multistrategy approach. In Machine Learning and Data Mining Methods and Applications.John Wiley, England.

....frequency [9] and an algorithm for discovering conceptual relations. A Multi Strategy Learning Result Set is used to support the complex task of ontology learning: It is possible to combine results from different learning methods, that have been applied to different sources. As described by [6] multi strategy learning architectures support balancing between advantages and disadvantages of different learning methods. The Ontology Engineering System OntoEdit supports the ontology engineer in semiautomatically adding newly discovered structures to the ontology. 1 In addition to core ....

R. Michalski and K. Kaufmann. Data mining and knowledge discovery: A review of issues and multistrategy approach. In Machine Learning and Data Mining Methods and Applications. John Wiley, England, 1998.

....2 VINLEN System 2. 1 An Overview Research on the VINLEN system grows out of our previous e#orts on the development of INLEN, an early system for integrating databases and machine learning and inference mechanisms for the purpose of multistrategy learning, data mining, and decision support [12,9]. INLEN included multiple learning and discovery operators, the high level knowledge generation language KGL1 [4,10] and an advisory system. It did not, however, integrate an actual database system, instead being constrained to relatively small tables located in reserved files. VINLEN is an ....

....structure of two rules for the decision High Blood Pressure is present Kenneth A. Kaufman and Ryszard S. Michalski users. Hence, visualization technology is very much in the spirit of inductive databases. We have developed visualization operators based on the method of diagrammatic visualization [9], and on association graphs [1,6] The latter method provides a novel approach to concept visualization. In it, the elements may represent attributes or high level concepts, with annotated links showing details of the relationships. Fig. 6, for example, presents an association graph built from ....

Michalski, R.S., Kaufman, K.A. (1998) Data Mining and Knowledge Discovery: A Review of Issues and a Multistrategy Approach. In Michalski, R.S., Bratko, I., Kubat, M. (eds.), Machine Learning and Data Mining: Methods and Applications, London, John Wiley & Sons, pp. 71-112

.... in the current INLEN system are the ability to learn different types of rules from examples, conceptual clustering and hierarchy generation, automatic selection of most relevant attributes, rule editing by an expert, and automatic application of the learned or acquired rules to new cases (Michalski and Kaufman, 1997). Important aspects of the INLEN approach that distinguish it from the most of existing data mining systems are that it employs a wide range of knowledge generation operators and is capable of knowledge intensive discovery. It allows a user to incorporate and utlilize various aspects of domain ....

....and interests to the system, so that the system can automatically perform desirable sequences of operators. To this end, we have initiated the development of KGL (Knowledge Generation Language) a meta level language for specifying knowledge discovery experiments using INLEN operators (Kaufman and Michalski, 1997). Specifically, the language allows the user to create plans of experiments and specify instructions for automatically guiding the system through sequences of steps and contingencies. The language is designed to support writing simple KGL programs that could perform very complex data mining and ....

[Article contains additional citation context not shown here]

Michalski, R.S. and Kaufman, K., "Data Mining and Knowledge Discovery: A Review of Issues and a Multistrategy Approach," Chapter in Michalski, R.S., Bratko, I. and Kubat, M. (eds.), Machine Learning and Data Mining: Methods and Applications, London, John Wiley & Sons, 1997 (to appear).

....discovery, and inference, based primarily on symbolic machine learning methods and techniques. INLEN integrates a range of knowledge generation operators, many of which represent different programs originally developed for use in stand alone machine learning applications (Michalski et al. 1992; Michalski and Kaufman 1998). The use of these operators allows one to discover general patterns, trends or exceptions in data that may not be apparent when only one type of strategy is applied. The results of applying diverse operators may also suggest a subsequent set of experiments that otherwise would not have been ....

....of a univariate AG representing statistical relationships in a medical domain is presented in Figure 7. The links thicknesses represent the magnitude of the weighted entropies, and the directions of the links indicate the higher conditional probability. A second study used the AQ18 program (Michalski 1998) to generate decision rules in this domain. The strongest rules for seven of the diseases are represented in the association graph shown in Figure 8. In this logical AG, the thickness of the links represents the informativeness level of a condition in the rule (based on the ratio of positive ....

Michalski, R.S. and Kaufman, K.A.. 1998. Data Mining and Knowledge Discovery: A review of Issues and a Multistrategy Approach. In Michalski, R.S., Bratko, I. and Kubat, M. eds. Machine Learning and Data Mining: Methods and Applications, 71-112. London: John Wiley & Sons.

....the results of its predecessors. These steps may involve the application of many diverse operators. In order to make these operators easily accessible to a data analyst; we have developed a system, INLEN, for multistrategy exploration of databases and decision support in real world applications (Michalski and Kaufman, 1998; Kaufman, Michalski and Kerschberg, 1991) INLEN offers a wide range of operators, such as those for determining conceptual (logic style) or statistical relationships among attributes, creating a characteristic or a discriminant description of classes of entities, optimizing an initial hypothesis ....

Michalski R.S., and Kaufman, K.A., Data Mining and Knowledge Discovery: A Review of Issues and a Multistrategy Approach, in Machine Learning and Data Mining: Methods and Applications, London, John Wiley & Sons, 1998.

No context found.

R. S. Michalski and K. A. Kaufman, "Data mining and knowledge discovery: A review of issues and a multistrategy approach," in Machine Learning and Data Mining: Methods and Applications, R. S. Michalski, I. Bratko, and M. Kubat, Eds. New York: Wiley, 1997.

No context found.

R. S. Michalski and K. A. Kaufman, "Data mining and knowledge discovery: A review of issues and a multistrategy approach," in Machine Learning and Data Mining: Methods and Applications, R. S. Michalski, I. Bratko, and M. Kubat, Eds. New York: Wiley, 1997.

No context found.

Michalski, R. S., and Kaufman, K. A. 1997. Data Mining and Knowledge Discovery: A Review of Issues and a Multistrategy Approach. In Machine Learning and Data Mining: Methods and Applications. Michalski, R. S.

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