Literature DB >> 21346963

Outlier Detection with One-Class SVMs: An Application to Melanoma Prognosis.

Stephan Dreiseitl1, Melanie Osl, Christian Scheibböck, Michael Binder.   

Abstract

BACKGROUND: Medical diagnosis and prognosis using machine learning methods is usually represented as a supervised classification problem, where a model is built to distinguish "normal" from "abnormal" cases. If cases are available from only one class, this approach is not feasible.
OBJECTIVE: To evaluate the performance of classification via outlier detection by one-class support vector machines (SVMs) as a means of identifying abnormal cases in the domain of melanoma prognosis.
METHODS: Empirical evaluation of one-class SVMs on a data set for predicting the presence or absence of metastases in melanoma patients, and comparison with regular SVMs and artificial neural networks.
RESULTS: One-class SVMs achieve an area under the ROC curve (AUC) of 0.71; two-class algorithms achieve AUCs between 0.5 and 0.84, depending on the available number of cases from the minority class.
CONCLUSION: One-class SVMs offer a viable alternative to two-class classification algorithms if class distribution is heavily imbalanced.

Entities:  

Mesh:

Year:  2010        PMID: 21346963      PMCID: PMC3041295     

Source DB:  PubMed          Journal:  AMIA Annu Symp Proc        ISSN: 1559-4076


  9 in total

1.  Estimating the support of a high-dimensional distribution.

Authors:  B Schölkopf; J C Platt; J Shawe-Taylor; A J Smola; R C Williamson
Journal:  Neural Comput       Date:  2001-07       Impact factor: 2.026

2.  Extraction of brain tumor from MR images using one-class support vector machine.

Authors:  J Zhou; K L Chan; V F Chong; S M Krishnan
Journal:  Conf Proc IEEE Eng Med Biol Soc       Date:  2005

3.  3D reconstruction of head MRI based on one class support vector machine with immune algorithm.

Authors:  Lei Wang; Guizhi Xu; Lei Guo; Xuena Liu; Shuo Yang
Journal:  Conf Proc IEEE Eng Med Biol Soc       Date:  2007

4.  The meaning and use of the area under a receiver operating characteristic (ROC) curve.

Authors:  J A Hanley; B J McNeil
Journal:  Radiology       Date:  1982-04       Impact factor: 11.105

5.  S-100beta and MIA in advanced melanoma in relation to prognostic factors.

Authors:  J M Auge; R Molina; X Filella; E Bosch; M Gonzalez Cao; S Puig; J Malvehy; T Castel; A M Ballesta
Journal:  Anticancer Res       Date:  2005 May-Jun       Impact factor: 2.480

6.  The revised American Joint Committee on Cancer staging system for melanoma.

Authors:  John A Thompson
Journal:  Semin Oncol       Date:  2002-08       Impact factor: 4.929

7.  Diagnostic value of melanoma inhibitory activity serum marker in the follow-up of patients with stage I or II cutaneous melanoma.

Authors:  Maja A Hofmann; Felix Gussmann; Andreas Fritsche; Susanne Biesold; Bernd Schicke; Ingeborg Küchler; Christiane Voit; Uwe Trefzer
Journal:  Melanoma Res       Date:  2009-02       Impact factor: 3.599

8.  Unsupervised spatiotemporal fMRI data analysis using support vector machines.

Authors:  Xiaomu Song; Alice M Wyrwicz
Journal:  Neuroimage       Date:  2009-03-31       Impact factor: 6.556

9.  Melanoma inhibiting activity protein (MIA), beta-2 microglobulin and lactate dehydrogenase (LDH) in metastatic melanoma.

Authors:  M González Cao; J M Auge; R Molina; R Martí; C Carrera; T Castel; R Vilella; C Conill; M Sánchez; J Malvehy; S Puig
Journal:  Anticancer Res       Date:  2007 Jan-Feb       Impact factor: 2.480

  9 in total
  2 in total

1.  Enhancement of hepatitis virus immunoassay outcome predictions in imbalanced routine pathology data by data balancing and feature selection before the application of support vector machines.

Authors:  Alice M Richardson; Brett A Lidbury
Journal:  BMC Med Inform Decis Mak       Date:  2017-08-14       Impact factor: 2.796

2.  A micro-XRT image analysis and machine learning methodology for the characterisation of multi-particulate capsule formulations.

Authors:  Frederik J S Doerr; Alastair J Florence
Journal:  Int J Pharm X       Date:  2020-01-15
  2 in total

北京卡尤迪生物科技股份有限公司 © 2022-2023.