Using the minimum description length principle to infer reduced ordered decision graphs

Oliveira, Arlindo L.; Sangiovanni‐Vincentelli, Alberto

doi:10.1007/bf00115299

Cited by 15 publications

(21 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The minimum description length principle (already investigated in the context of decision diagrams in [17]) might become a valuable tool for increasing the generalization abilities of decision diagrams, while at the same time reducing their size.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Extracting Rules From Neural Networks as Decision Diagrams

Chorowski

Żurada

2011

IEEE Trans. Neural Netw.

View full text Add to dashboard Cite

Abstract-Rule extraction from neural networks solves two fundamental problems: it gives insight into the logic behind the network and, in many cases, it improves the network's ability to generalize the acquired knowledge. This article presents a novel, eclectic approach to rule extraction from neural networks, named LORE, suited for multilayer perceptron networks with discrete (logical or categorical) inputs. The extracted rules mimic network behavior on training set and relax this condition on the remaining input space.First, a multilayer perceptron network is trained under standard regime. It is then transformed into an equivalent form, returning the same numerical result as the original network, yet being able to produce rules generalizing the network output for cases similar to a given input. The partial rules extracted for every training set sample are then merged to form a decision diagram from which logic rules can be extracted.A rule format explicitly separating subsets of inputs for which an answer is known from those with an undetermined answer is presented. A special data structure, the decision diagram, allowing efficient partial rule merging is introduced. With regard to rules' complexity and generalization abilities, LORE gives results comparable with those reported previously. An algorithm transforming decision diagrams into interpretable boolean expressions is described. Experimental running times of rule extraction are proportional to the network's training time.Index Terms-Feedforward neural networks, rule extraction, logic rules, true false unknown logic, decision diagrams.

show abstract

Section: Discussionmentioning

confidence: 99%

“…A method for top-down induction of RODDs is presented by Kohavi in [16]. In [17], Oliveira and Sangiovanni-Vincentelli show an interesting study of a pruning algorithm used for generalization. The suitability of RODDs for visualization of simple decision rules is studied in [18].…”

Section: Decision Diagramsmentioning

confidence: 99%

Extracting Rules From Neural Networks as Decision Diagrams

Chorowski

Żurada

2011

IEEE Trans. Neural Netw.

View full text Add to dashboard Cite

show abstract

“…Vapnik (1999) provided a formal proof to justify the principle for classification problems. Although much work remains to be done on effective constructive procedures under the principle (Cherkassky and Mulier 1998), it has been applied for machine learning (Yamanishi 1992;Zemel 1993;Pfahringer 1995;Rissanen and Yu 1996), model selection (Hansen and Yu 1998), coding decision trees and graphs (Wallace and Patrick 1993;Oliveira and Sangiovanni-Vincentelli 1995), and even for classification of protein structure (Edgoose, Allison, and Dowe 1998).…”

Section: Introductionmentioning

confidence: 99%

Inductive Inference by Using Information Compression

Choi¹

2003

Computational Intelligence

View full text Add to dashboard Cite

Inductive inference is of central importance to all scientific inquiries. Automating the process of inductive inference is the major concern of machine learning researchers. This article proposes inductive inference techniques to address three inductive problems: (1) how to automatically construct a general description, a model, or a theory to describe a sequence of observations or experimental data, (2) how to modify an existing model to account for new observations, and (3) how to handle the situation where the new observations are not consistent with the existing models. The techniques proposed in this article implement the inductive principle called the minimum descriptive length principle and relate to Kolmogorov complexity and Occam's razor. They employ finite state machines as models to describe sequences of observations and measure the descriptive complexity by measuring the number of states. They can be used to draw inference from sequences of observations where one observation may depend on previous observations. Thus, they can be applied to time series prediction problems and to one-to-one mapping problems. They are implemented to form an automated inductive machine.

show abstract

“…Primary focuses of these researches are controlling tree size (Kim and Koehler, 1994;Auer et al, 1995; Kalkanis, 1993; 18 Distribution A: Approved for public release; distribution is unlimited. 88ABW-2011-0727, 23 Feb 2011 Esposito et al, 1997;Smyth et al, 1995), modifying test space (Fayyad and Irani, 1992; Utgoff and Brodley, 1990;Heath et al, 1993b;Zheng, 1995), modifying test search (Quinlan, 1986;Dietterich et al, 1996; Utgoff and Clouse, 1996;Pazzani et al, 1994), database restrictions (Wirth and Catlett, 1988;John, 1995) and alternative data structures (Oliver, 1993;Oliveira and Sangiovanni-Vincentelli, 1995; Kohavi and Li, 1995). …”

mentioning

confidence: 99%

An Intelligent Decision Support System for Workforce Forecast

Wan¹,

Chen²,

Kuriger³

2011

View full text Add to dashboard Cite

Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing this collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden to Department of Defense, Washington Headquarters Services, Directorate for Information , 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to any penalty for failing to comply with a collection of information if it does not display a currently valid OMB control number. PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS. REPORT DATE (DD-MM-YYYY) 05-01-2011 REPORT TYPE 7184D214 PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION REPORT NUMBERThe University of Texas at San Antonio One UTSA Circle San Antonio TX 78249 SPONSORING / MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR'S ACRONYM(S)Air SPONSOR/MONITOR'S REPORT NUMBER(S)AFRL-RH-WP-TR-2011-0019 DISTRIBUTION / AVAILABILITY STATEMENTDistribution A: Approved for public release; distributed is unlimited. SUPPLEMENTARY NOTES88ABW/PA cleared on 23 Feb 2011, 88ABW-2011-0727. ABSTRACTThe main objective of this project is to enhance the effectiveness of workforce forecast and deployment through innovative approaches of artificial intelligence. The research included a final document of conventional and innovative methods of workforce forecasting and a decision support software program incorporating advanced artificial intelligence techniques for workforce forecasting. SUBJECT TERMS ABSTRACTThis project studies workforce forecasting in two main aspects: (1) an extensive review of the existing methodologies and techniques and (2) an effort to develop a decision support system with models and software programming. The main objective of this project is to enhance the effectiveness of workforce forecasting and deployment through innovative approaches of artificial intelligence. The deliverables include a summary report of conventional and innovative methods of workforce forecasting (Part I) and a decision support software program using artificial intelligence techniques for workforce forecasting (Part II).Part I, provided a thorough literature review of fundamental research and practices of demand and supply forecasting techniques for workforce analysis. Over 300 relevant literatures have been identified and reviewed by the project team, and 289 of them are covered in this report.Following is a summary of the main contents and contributions of Part I:• This report provides an extensive review of the fundamental knowledge, applications, guidelines, and scope of use of various workforce forecasting methods.• A decision tree has been proposed for selecting an appropriate forecasting ...

show abstract

Using the minimum description length principle to infer reduced ordered decision graphs

Cited by 15 publications

References 27 publications

Extracting Rules From Neural Networks as Decision Diagrams

Extracting Rules From Neural Networks as Decision Diagrams

Inductive Inference by Using Information Compression

An Intelligent Decision Support System for Workforce Forecast

Contact Info

Product

Resources

About