Extracting easily interpreted diagnostic rules

Porębski, Sebastian; Straszecka, Ewa

doi:10.1016/j.ins.2017.10.034

Cited by 13 publications

(12 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The medical literature abounds with measures or indicators for the statistical analysis of screening and diagnostic (confirmatory) tests, which are conducted on asymptomatic subjects and on subjects having symptoms (signs), respectively. There are many informative tutorials on the topic, including the contributions of Alonzo and Pepe (1999), Glas et al (2003), Hawkins (2005), Fawcett (2006), Zhou et al (2009), Powers (2011), Lewis and Torgerson (2012), Broemeling (2011), Leeflang (2014), Chughtai et al (2015), Kent and Hancock (2016), Baveja and Aggarwal (2017), Porebski and Straszecka, (2018). Fig.…”

Section: Appendix A: Prominent Measures or Indicators In The Statistical Analysis Of Screening And Diagnostic Testsmentioning

confidence: 99%

Karnaugh-Map Utility in Medical Studies: The Case of Fetal Malnutrition

Rushdi¹,

Rushdi²

2018

Int J Math, Eng, Manag Sci

View full text Add to dashboard Cite

This paper advocate and demonstrates the utility of the Karnaugh map, as a pictorial manual tool of Boolean algebra, in the exploration of medical problems as exemplified herein by the problem of Fetal Malnutrition (FM). The paper briefly introduces the FM problem, and specifies four metrics or tests used frequently in its study. Clinical data collected about these metrics (as continuous variables or dichotomized versions thereof) are conventionally treated via statistical methods. The Karnaugh map serves as a convenient way for aggregating the set of clinical data available into a pseudo-Boolean function. The map can be used to produce a two-by-two contingency matrix (confusion matrix or frequency matrix) that relates an assessed test or metric to a reference or standard one. Each of these two metrics can be any of the map variables or a function of some or all of these variables. While the map serves in this capacity as a supplement or aid to statistical methods, it is also useful for certain non-statistical methods (specifically Boolean ones). The paper shows how the map entries can be dichotomized via an appropriate threshold for use in Boolean Analysis (BA), which can be conducted despite the lack of a gold standard. The map also implements Qualitative Comparative Analysis (QCA) for the given clinical data. The map variable-handling capability does not pose as a shortcoming for either BA or QCA, since the number of variables involved (not only herein but in other typical medical problems as well) is relatively small. The concepts and methods introduced herein are demonstrated through application to the aforementioned set of clinical data for the FM problem, and can be extended to a wide variety of medical problems.

show abstract

Section: Appendix A: Prominent Measures or Indicators In The Statistical Analysis Of Screening And Diagnostic Testsmentioning

confidence: 99%

Karnaugh-Map Utility in Medical Studies: The Case of Fetal Malnutrition

Rushdi¹,

Rushdi²

2018

Int J Math, Eng, Manag Sci

View full text Add to dashboard Cite

show abstract

“…Future studies can extend the belief intervals combination rule in the following directions. There are numerous aggregation operators in the fuzzy sets domain, in this article, we only select the IFWA operator to construct the belief intervals combination rule and some other aggregation operators are also expected to be used in the combination rules of DST. Moreover, the DST has been generated with such focal sets as multigranulation spaces, discrete belief, fuzzy sets, multidimensional data, belief with missing data values, and belief with various types of belief . It will be interesting to extend the belief intervals combination rule these generalized DST.…”

Section: Discussionmentioning

confidence: 99%

“…Moreover, the DST has been generated with such focal sets as multigranulation spaces, discrete belief, fuzzy sets, multidimensional data, belief with missing data values, and belief with various types of belief . It will be interesting to extend the belief intervals combination rule these generalized DST.…”

Section: Discussionmentioning

confidence: 99%

Belief intervals aggregation

Sun¹,

Guan²,

Xiao³

et al. 2018

Int. J. Intell. Syst.

View full text Add to dashboard Cite

The combination rule is the core of Dempster-Shafer theory (DST), and there is no uniform aggregation rule adapting to all conditions. The construction of such a ruleis still an open and hot topic. In this article, we focus on this point. We focus on the belief interval, which is made up of belief function and plausibility function, in the form of [belief function, plausibility function], instead of basic belief assignment, to represent the DST. We aim at exploring a belief interval combination rule as the combination rule of DST. To do this, we contrast the belief interval with the intuitionistic fuzzy sets and construct the belief interval combination rule based on the intuitionistic fuzzy weighted averaging (IFWA) operator, which opens the door for DST combination rules to the aggregation operator perspective. Further, we find that directly using the IFWA operator as the belief interval combination rule poses three problems: the "'one' veto problem," converge easily close to [1,1] and the belief function from the belief interval combination rule is not normalized. To solve these problems, we add a normalization process in the belief interval combination rule to modify it, which can address all these three problems well.We also set up a series of examples to illustrate the belief interval combination rule, including a multisensor fusion scenario to compare it with some of the existing rules, which shows its superiorities in better stability and lower operational load. K E Y W O R D S aggregation operator, belief functions, belief intervals, combination rule, plausibility function Int J Intell Syst. 2018;2425-2447.wileyonlinelibrary.com/journal/int

show abstract

“…How to elicit symptom-diagnosis rules from medical data subject to uncertainty or ambiguity, and how to make a synthesized diagnosis when a Q&A thread is covered by multiple rules? Some promising results on mining text subject to uncertainty have been obtained (Boegl et al, 2004;Chinnnaswamy & Srinivasan, 2018;Malmir et al, 2017;Porebski & Straszecka, 2018;Stoklasa et al, 2017;Sun et al, 2011;Tsipouras et al, 2008;Wang et al, 2016;Wang & Lee, 2011), but with regards to eliciting uncertain information and combining rules, there is still much to explore. Different from the above-mentioned fuzzy expert systems, the knowledge bases in this study are generated within belief rule-based structures so that the diagnostic consequent could reflect slight changes of linguistic variables occurring in the symptom-related text.…”

Section: How To Distinguish Diseases Sharing Similar Symptoms?mentioning

confidence: 99%

Eliciting symptom‐diagnosis knowledge from online medical Q&A

Huang

2021

Expert Systems

View full text Add to dashboard Cite

With the objective to automatically detect diseases from symptoms in free-text data, a methodology to extract symptom-diagnosis knowledge from online medical textual data in Q&A domain is proposed in this paper: (1) a term frequency-inverse document frequency and PRECISION method is adopted to retrieve symptom words from unstructured text; (2) a variable precision rough set based genetic algorithm is applied to reduce redundant symptom words, and a rough set based rule is utilized for adding discriminative symptom words assisting to discriminate diseases sharing similar symptoms; (3) by employing fuzzy linguistic variables to express the risk level of disease or severity level of symptoms, a knowledge base with fuzzy belief structure is generated. Using data extracted from a Chinese medical Q&A forum for training and testing, some classical gastrointestinal diseases serve as a case study to evaluate the efficiency of the proposed methodology. Subsequently performance comparisons are made between the proposed methodology and some other classifiers, such as the decision tree algorithms including ID3 and J45, and the Bayesian network classifier.The comparative results demonstrate that the proposed methodology outperforms the decision tree algorithms and the Bayesian network classifier.

show abstract

Extracting easily interpreted diagnostic rules

Cited by 13 publications

References 37 publications

Karnaugh-Map Utility in Medical Studies: The Case of Fetal Malnutrition

Karnaugh-Map Utility in Medical Studies: The Case of Fetal Malnutrition

Belief intervals aggregation

Eliciting symptom‐diagnosis knowledge from online medical Q&A

Contact Info

Product

Resources

About