LERS-A System for Learning from Examples Based on Rough Sets

Grzymała-Busse, Jerzy W.

doi:10.1007/978-94-015-7975-9_1

Cited by 499 publications

(115 citation statements)

References 32 publications

(7 reference statements)

Supporting

Mentioning

115

Contrasting

Order By: Relevance

“…In our research, the main data mining tool was LERS (Learning from Examples based on Rough Sets), developed at the University of Kansas (Grzymala-Busse, 1992). LERS has proven its applicability having been used for years by NASA Johnson Space Center (Automation and Robotics Division), as a tool to develop expert systems of the type most likely to be used in medical decision-making on board the International Space Station.…”

Section: Rule Inductionmentioning

confidence: 99%

An Approach to Imbalanced Data Sets Based on Changing Rule Strength

Grzymała-Busse

Goodwin

Grzymala-Busse

et al. 2004

Rough-Neural Computing

Self Cite

View full text Add to dashboard Cite

This paper describes experiments with a challenging data set describing preterm births. The data set, collected at the Duke University Medical Center, was large and, at the same time, many attribute values were missing. However, the main problem was that only 20.7% of the total number of cases represented the important preterm birth class. Thus the data set was imbalanced. For comparison, we include results of experiments on another imbalanced data set, the wellknown breast cancer data set. Our approach to dealing with this imbalanced data set was to induce a rule set using our standard procedure: the LEM2 algorithm of the LERS rule induction system and then increase the rule strength for all rules describing preterm births by multiplying all such rule strengths by the same number called a strength multiplier. The rules strength for any rule describing the majority class, fullterm birth, remained unchanged. The optimal strength multiplier was determined experimentally using our optimality criterion: the maximum of the sum of sensitivity and specificity.

show abstract

Section: Rule Inductionmentioning

confidence: 99%

An Approach to Imbalanced Data Sets Based on Changing Rule Strength

Grzymała-Busse

Goodwin

Grzymala-Busse

et al. 2004

Rough-Neural Computing

Self Cite

View full text Add to dashboard Cite

show abstract

“…This specialization process is repeated until a selected acceptance criterion has been fulfilled, e.g. the current condition part does not cover any of the negative examples (e.g., see the description of AQ [26]or LEM2 [10]). …”

Section: Induction Of Minimal Sets Of Rulesmentioning

confidence: 99%

“…The most radical solution is to produce so called exhaustive set of rules, which contains all (often discriminant only) rules that can be induced on the basis of positive examples of the class. Examples of such approach include the dropping condition technique, described in [10] or Boolean reasoning approach to look for local object reducts [32] -specific for rough set theory. However, time complexity for the second choice is exponential and using this kind of approach may be not practical for larger input data filesso approximate algorithms are employed.…”

Section: Induction Of Non-minimal Sets Of Rulesmentioning

confidence: 99%

“…To evaluate the usefulness of our approach we experimentally compared it to a baseline classifier using minimal set of rules generated by LEM2 [10]. Moreover, we considered a variant of such LEM2 classifier with a modified classification strategy expanded with a support multiplier (such modification is particularly suited to deal with imbalanced data -see its description in Section 4).…”

Section: Experimental Evaluationmentioning

confidence: 99%

See 1 more Smart Citation

Extending Rule-Based Classifiers to Improve Recognition of Imbalanced Classes

Stefanowski

Wilk

2009

Advances in Data Management

View full text Add to dashboard Cite

Abstract. This papers deals with inducing rule-based classifiers from imbalanced data, where one class (a minority class) is under-represented in comparison to the remaining classes (majority classes). We discuss reasons for bias of standard classifiers toward recognition of examples from majority classes and misclassifcation of the minority class. To avoid limitations of sequential covering approaches, we present a new approach to improve sensitivity of the rule based classifier. It includes a modification the structure of sets of rules, where for majority classes minimal sets of rules are still induced while the rule set for the minority class is generated by the algorithm, called EXPLORE. This algorithm produces rules being more general and supported by more learning examples than rules from the minimal set. The usefulness of the new approach is verified in a comparative experiments on several imbalanced data sets.

show abstract

“…The LERS (Learning from Examples based on Rough Sets) program utilized to generate the rules during this research is the descendent of several generations of artificial intelligence programs that learn from experience (Grzymala-Busse, 1992;p. 2, 1988).…”

Section: Rule Induction the Lers Approach To Rule Inductionmentioning

confidence: 99%

Global temperature stability by rule induction: An interdisciplinary bridge

Gunn

Grzymała-Busse

1994

Hum Ecol

Self Cite

View full text Add to dashboard Cite

Abstract:Rules incorporating influences on global temperature, an estimate of radiation balance, were induced from astronomical, geophysical, and anthropogenic variables. During periods of intermediate global temperatures (generally like the present century), the influences assume canceling roles; influences cancel the effects of extreme states potentially imposed by other influences because they are, in aggregate, most likely to be assuming opposite values. This imparts an overall stability to the global temperature. To achieve cold or hot global temperature, influences assume reinforcing roles. CO 2 is an active influence on global temperature. By virtue of its constancy in the atmosphere, it can be expected to sponsor frequent hot years in combination with the other influences as they cycle through their periods. If measures were implemented to maintain warm or cool global temperatures, it could retain the status quo of present global agricultural regions. They are probably more productive than hot world regions would be because of narrow storm tracks. KEY WORDS: global climate; artificial intelligence; energy balance; rough sets. Article:INTRODUCTION The articles in this issue discuss variations in the sources of energy that modify the earth's radiation budget. They are features of the earth's orbit, atmosphere, oceans, and geosphere that modulate the amount of energy admitted to, and retained within, the oceans and atmosphere. In this article we will examine the interaction of these influences within the modern atmosphere through annual global average temperature, an estimate of radiation budget. Because of the concern generated during the 1980s by numerous record high global temperature years, and questions about future change (Houghton et al., 1990), we direct our attention to the mechanism of global temperature stability.The physical sciences have reasonably been concerned with the trendier aspects of global climate change. In the life sciences, occasional intimations of concern for interannual climate stability have begun to appear in the literature. Efforts to understand these life system effects can be expected to yield more easily to examination and explanation as year-by-year stability indices spanning thousands of years become more common. Thompson et al. (this issue) discuss the considerable range of information present in ice cores on global conditions. Through tree rings, Dean et al. (1985) provide a year-by-year index of regional climate stability in the southwestern United States spanning thousands of years. We make use of the worldwide and highly accurate measures of atmospheric temperature and related phenomena during the last decades to study temperature change processes. Close scrutiny of these data provides insights into the changes recorded in the long records. Results may apply across the sciences as historic and prehistoric climate records are drawn into the debate over anthropogenic contributions to global change. Rightly so, as the unforseen problems of global change, coupled with u...

show abstract

LERS-A System for Learning from Examples Based on Rough Sets

Cited by 499 publications

References 32 publications

An Approach to Imbalanced Data Sets Based on Changing Rule Strength

An Approach to Imbalanced Data Sets Based on Changing Rule Strength

Extending Rule-Based Classifiers to Improve Recognition of Imbalanced Classes

Global temperature stability by rule induction: An interdisciplinary bridge

Contact Info

Product

Resources

About