Abstract
The aim of many microarray experiments is to build discriminatory diagnosis and prognosis models. A large number of supervised methods have been proposed in literature for microarray-based classification. Model comparison, which is based on the classification error estimation, is a critical issue. Previous studies have shown that error estimation is unreliable in high-dimensional small-sample settings. This leads naturally to questioning the validity of classificationrule comparison approaches being used in the literature. In this paper we present a brief review of the different comparison methods used in bioinformatics. Then, we test these methods on a set of simulations based on both synthetic and real data. These simulations include different feature-label distributions, classification rules, error estimators and variance estimators. The results show that none of these methods can provide reliable comparison across a wide spectrum of feature-label distributions and classification rules.
Keywords: Microarray classification, error estimation, classifier comparison, variance study
Current Bioinformatics
Title: On the Comparison of Classifiers for Microarray Data
Volume: 5 Issue: 1
Author(s): Blaise Hanczar and Edward R. Dougherty
Affiliation:
Keywords: Microarray classification, error estimation, classifier comparison, variance study
Abstract: The aim of many microarray experiments is to build discriminatory diagnosis and prognosis models. A large number of supervised methods have been proposed in literature for microarray-based classification. Model comparison, which is based on the classification error estimation, is a critical issue. Previous studies have shown that error estimation is unreliable in high-dimensional small-sample settings. This leads naturally to questioning the validity of classificationrule comparison approaches being used in the literature. In this paper we present a brief review of the different comparison methods used in bioinformatics. Then, we test these methods on a set of simulations based on both synthetic and real data. These simulations include different feature-label distributions, classification rules, error estimators and variance estimators. The results show that none of these methods can provide reliable comparison across a wide spectrum of feature-label distributions and classification rules.
Export Options
About this article
Cite this article as:
Hanczar Blaise and Dougherty R. Edward, On the Comparison of Classifiers for Microarray Data, Current Bioinformatics 2010; 5 (1) . https://dx.doi.org/10.2174/157489310790596376
DOI https://dx.doi.org/10.2174/157489310790596376 |
Print ISSN 1574-8936 |
Publisher Name Bentham Science Publisher |
Online ISSN 2212-392X |
- Author Guidelines
- Graphical Abstracts
- Fabricating and Stating False Information
- Research Misconduct
- Post Publication Discussions and Corrections
- Publishing Ethics and Rectitude
- Increase Visibility of Your Article
- Archiving Policies
- Peer Review Workflow
- Order Your Article Before Print
- Promote Your Article
- Manuscript Transfer Facility
- Editorial Policies
- Allegations from Whistleblowers
Related Articles
-
Targeting Bcl-2 in CLL
Current Medicinal Chemistry Nucleic Acid-Based Aptamers: Applications, Development and Clinical Trials
Current Medicinal Chemistry MicroRNAs in Human Diseases
Recent Patents on DNA & Gene Sequences The Effect of Lipoic Acid on Macro and Trace Metal Levels in Living Tissues Exposed to Oxidative Stress
Anti-Cancer Agents in Medicinal Chemistry Castration-resistant Prostate Cancer: Novel Therapeutics Pre- or Post- Taxane Administration
Current Cancer Drug Targets Peptide-Drug Conjugate: A Novel Drug Design Approach
Current Medicinal Chemistry Aptamers Against Cell Surface Receptors: Selection, Modification and Application
Current Medicinal Chemistry Targeting Tumor-Associated Antigens to the MHC Class I Presentation Pathway
Endocrine, Metabolic & Immune Disorders - Drug Targets Is there a Possible Single Mediator in Modulating Neuroendocrine–thymus Interaction in Ageing?
Current Aging Science Assays for Identification of Hsp90 Inhibitors and Biochemical Methods for Discriminating their Mechanism of Action
Current Topics in Medicinal Chemistry β2-AR-HIF-1α: A Novel Regulatory Axis for Stress-Induced Pancreatic Tumor Growth and Angiogenesis
Current Molecular Medicine Liposomes as siRNA Delivery Vectors
Current Drug Metabolism Calcium-calpain Dependent Pathways Regulate Vesiculation in Malignant Breast Cells
Current Cancer Drug Targets Silencing Human Cancer: Identification and Uses of MicroRNAs
Recent Patents on Anti-Cancer Drug Discovery Bladder Cancer and Stem Cells
Current Signal Transduction Therapy Insulin-like Growth Factor-1 Receptor (IGF-1R) Kinase Inhibitors in Cancer Therapy: Advances and Perspectives
Current Pharmaceutical Design Epigenetic control of cardiovascular health by nutritional polyphenols involves multiple chromatin-modifying writer-reader-eraser proteins
Current Topics in Medicinal Chemistry Aldehyde Dehydrogenase as a Marker for Stem Cells
Current Stem Cell Research & Therapy Editorial (Thematic Issue: Chemoinformatics Models for Pharmaceutical Design, Part 2)
Current Pharmaceutical Design New Approaches for the Synthesis of Fused Thiophene, Pyrazole, Pyran and Pyridine Derivatives with Anti-Proliferative Together with c-Met Kinase and Prostate Cancer Cell Inhibitions
Anti-Cancer Agents in Medicinal Chemistry