Capped K-NN Editing in Definition Lacking Environments

Stranieri, Andrew; Yatsko, Andrew; Golden, Isaac; Mammadov, Musa; И, Багиров А

doi:10.13176/11.465

Cited by 2 publications

(7 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The class imbalance (also 'skew') [3,4,[18][19][20] is more profound in the DM problem than the two others. The prevalence of DM is roughly half of that for CVD or HT.…”

Section: Resultsmentioning

confidence: 99%

“…This is on the part of class imbalance. [3,4,[18][19][20] So, the objective presently was to prop the minimum of the two instead of the balanced accuracy.…”

Section: Resultsmentioning

confidence: 99%

“…The second requires another pass through the data and can be considered a general method outside the competence of space metric. [3,4] Cost and Salzberg noted that their approach lends itself easily to parallel computations, therefore naming it the Parallel Exemplar-Based Learning System (PE-BLS). [2] They however acknowledged that this may appear overreaching and would equally apply to the original method of Stanfill and Waltz.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Weighting features by the value displacement rebound

Yatsko

2020

AIR

Self Cite

View full text Add to dashboard Cite

Learning from examples draws on similarity, a concept which formalisation leads to the notion of instance space. Continuous spaces are easier to embrace since, unlike discrete, they often can be seen as hyper-constructs of 3D. Unsurprisingly, the instance-based learning methods are more developed for continuous domains than for discrete ones. The value difference metric (VDM) is one of the few examples of metrics for discrete spaces. Mixed reports about utility of VDM exist. In this paper VDM is compared with another approach where data features are weighted by the Information Gain. Some vulnerabilities of VDM are identified. A weighting method, nothing like VDM, although inspired by the former, is proposed. The results are in favour of the new weighting scheme with illustration of utility for health diagnostics.

show abstract

“…The class imbalance (also 'skew') [3,4,[18][19][20] is more profound in the DM problem than the two others. The prevalence of DM is roughly half of that for CVD or HT.…”

Section: Resultsmentioning

confidence: 99%

“…This is on the part of class imbalance. [3,4,[18][19][20] So, the objective presently was to prop the minimum of the two instead of the balanced accuracy.…”

Section: Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Weighting features by the value displacement rebound

Yatsko

2020

AIR

Self Cite

View full text Add to dashboard Cite

show abstract

“…Gonzalez-Reyna et al [34] in their traffic sign recognition problem apply techniques that, instead of under-sampling of the majority class, over-sample the minority class or do both, particularly the technique by Chawla et al [35] We proposed solving of the imbalanced data problem by class noise reduction in ref. [10] Despite being very class imbalanced, the Secom example has many instances and attributes, posing a challenge for quick evaluation. Therefore, only half of the dataset in Table 1 was used.…”

Section: Diagnostic Domainsmentioning

confidence: 99%

“…We discuss the intrusive measures in ref. [10] In that regard, using the paradigm of attribute space facilitates reckoning; however, applying the notion is not straightforward in all-nominal / discrete data domains. Techniques, akin developed in ref.…”

Section: Introductionmentioning

confidence: 99%

Diagnostic with incomplete nominal/discrete data

Jelinek¹,

Yatsko²,

Stranieri³

et al. 2015

AIR

Self Cite

View full text Add to dashboard Cite

Missing values may be present in data without undermining its use for diagnostic / classification purposes but compromise application of readily available software. Surrogate entries can remedy the situation, although the outcome is generally unknown. Discretization of continuous attributes renders all data nominal and is helpful in dealing with missing values; particularly, no special handling is required for different attribute types. A number of classifiers exist or can be reformulated for this representation. Some classifiers can be reinvented as data completion methods. In this work the Decision Tree, Nearest Neighbour, and Naive Bayesian methods are demonstrated to have the required aptness. An approach is implemented whereby the entered missing values are not necessarily a close match of the true data; however, they intend to cause the least hindrance for classification. The proposed techniques find their application particularly in medical diagnostics. Where clinical data represents a number of related conditions, taking Cartesian product of class values of the underlying sub-problems allows narrowing down of the selection of missing value substitutes. Real-world data examples, some publically available, are enlisted for testing. The proposed and benchmark methods are compared by classifying the data before and after missing value imputation, indicating a significant improvement.

show abstract

Capped K-NN Editing in Definition Lacking Environments

Cited by 2 publications

References 0 publications

Weighting features by the value displacement rebound

Weighting features by the value displacement rebound

Diagnostic with incomplete nominal/discrete data

Contact Info

Product

Resources

About