Advertisement

Algorithm of Multidimensional Analysis of Main Features of PCA with Blurry Observation of Facility Features Detection of Carcinoma Cells Multiple Myeloma

Conference paper
Part of the Advances in Intelligent Systems and Computing book series (AISC, volume 977)

Abstract

The article is mainly focused on the description of the PCA main component algorithm with fuzzy observation of object features. The author focuses on the application of this method to the reduction of similar traits, aiming at the diagnosis of not only the existence of cancerous rhinoceros, but also the assessment of the direction of its exposure and the degree of aggressiveness. An advantage of the developed algorithm is the ability to combine basic patient results with morphology or advanced indicators of tumor markers with microscopic or X-ray image. The occurrence is a possibility of the development of a group of people. Thanks to the fuzzy observation of the object. For this purpose, the author’s points are the correlations. This gives a positive answer, where redundancy can indicate the same class. The analysis of the principal components does not exhaust the subject of research. Multidimensional exploration techniques that can improve the computerized medical diagnostics. The proposed solution can be used for unbalanced data. In the course of further research, it is possible to use the data imputation method. The method developed for selection of the traits of the subject is an original approach. It allows us to choose traits when data are either unconfirmed or incomplete. When creating statistical models for various medical institutions, the author encountered many classification problems. For doctors, the use of traditional statistical methods did not give satisfactory results. Very often the selection of traits was not very precise and could raise doubts as to its accuracy. The use of algorithms of fuzzy logic in PCA can be the foundation for a more accurate trait selection by combining different quantitative data. In the case of selecting cancer traits, it allows us to combine image data with test data. That way, it is possible to get additional information about the direction of cancer cell growth.

Keywords

Fuzzy sets Main components analysis Recognition Multiple myeloma 

References

  1. 1.
    Webb AR, Copsey KD (2011) Statistical pattern recognition. Wiley.  https://doi.org/10.1109/tpami.2015.2491929CrossRefGoogle Scholar
  2. 2.
    Krawczyk B, Woźniak M (2014) Hypertension type classification using hierarchical ensemble of one-class classifiers for imbalanced data. Springer, pp 341–349.  https://doi.org/10.1109/tpami.2015.2491929CrossRefGoogle Scholar
  3. 3.
    Krawczyk B, Woźniak M (2015) Combining nearest neighbour classifiers based on small subsamples for big data analytics. CYBCONF, pp 24–26.  https://doi.org/10.1109/tpami.2015.2491929CrossRefGoogle Scholar
  4. 4.
    Burduk R (2017) The adaboost algorithm with linear modification of the weights. In: 9th international conference, Bydgoszcz, Poland.  https://doi.org/10.1109/tpami.2015.2491929CrossRefGoogle Scholar
  5. 5.
    Han J, Kamber M, Pei J (2012) Concept and techniques, 3rd edn. Elsevier.  https://doi.org/10.1109/tpami.2015.2491929CrossRefGoogle Scholar
  6. 6.
    Turk MA, Pentland AP (1991) Face recognition using eigenface. In: Conference on computer vision and pattern recognition, pp 586–591.  https://doi.org/10.1109/tpami.2015.2491929CrossRefGoogle Scholar
  7. 7.
    Trajdos P, Kurzynski M (2015) An extension of multi-label binary relevance models based on randomized reference classifier and local fuzzy confusion matrix. In: 16th international conference.  https://doi.org/10.1109/tpami.2015.2491929CrossRefGoogle Scholar
  8. 8.
    Sobczak W, Malina W (1985) Metody selekcji i redukcji informacji. WNT.  https://doi.org/10.1109/tpami.2015.2491929CrossRefGoogle Scholar

Copyright information

© Springer Nature Switzerland AG 2020

Authors and Affiliations

  1. 1.Department of Systems and Computer NetworksWroclaw University of TechnologyWroclawPoland

Personalised recommendations