Transfer Learning for Molecular Cancer classification using Deep Neural Networks

Rahul Kumar Sevakula, Vikas Singh, Nishchal K. Verma, Chandan Kumar, Yan Cui

Research output: Contribution to journalArticle

7 Citations (Scopus)

Abstract

The emergence of deep learning has impacted numerous machine learning based applications and research. The reason for its success lies in two main advantages: 1) it provides the ability to learn very complex non-linear relationships between features and 2) it allows one to leverage information from unlabeled data that does not belong to the problem being handled. This paper presents a transfer learning procedure for cancer classification, which uses feature selection and normalization techniques in conjunction with stacked sparse auto-encoders on gene expression data. While classifying any two tumor types, data of other tumor types were used in unsupervised manner to improve the feature representation. The performance of our algorithm was tested on 36 two-class benchmark datasets from the GEMLeR repository. On performing statistical tests, it is clearly ascertained that our algorithm statistically outperforms several generally used cancer classification approaches. The deep learning based molecular disease classification can be used to guide decisions made on the diagnosis and treatment of diseases, and therefore may have important applications in precision medicine.

Original languageEnglish (US)
JournalIEEE/ACM Transactions on Computational Biology and Bioinformatics
DOIs
StateAccepted/In press - Apr 3 2018

Fingerprint

Cancer Classification
Transfer Learning
Tumor
Neural Networks
Tumors
Statistical test
Gene Expression Data
Encoder
Leverage
Repository
Feature Selection
Medicine
Normalization
Neoplasms
Machine Learning
Statistical tests
Learning
Benchmark
Gene expression
Benchmarking

All Science Journal Classification (ASJC) codes

  • Biotechnology
  • Genetics
  • Applied Mathematics

Cite this

Transfer Learning for Molecular Cancer classification using Deep Neural Networks. / Sevakula, Rahul Kumar; Singh, Vikas; Verma, Nishchal K.; Kumar, Chandan; Cui, Yan.

In: IEEE/ACM Transactions on Computational Biology and Bioinformatics, 03.04.2018.

Research output: Contribution to journalArticle

@article{f7b042cc9d0a4dd6a6e364b46dc56aa1,
title = "Transfer Learning for Molecular Cancer classification using Deep Neural Networks",
abstract = "The emergence of deep learning has impacted numerous machine learning based applications and research. The reason for its success lies in two main advantages: 1) it provides the ability to learn very complex non-linear relationships between features and 2) it allows one to leverage information from unlabeled data that does not belong to the problem being handled. This paper presents a transfer learning procedure for cancer classification, which uses feature selection and normalization techniques in conjunction with stacked sparse auto-encoders on gene expression data. While classifying any two tumor types, data of other tumor types were used in unsupervised manner to improve the feature representation. The performance of our algorithm was tested on 36 two-class benchmark datasets from the GEMLeR repository. On performing statistical tests, it is clearly ascertained that our algorithm statistically outperforms several generally used cancer classification approaches. The deep learning based molecular disease classification can be used to guide decisions made on the diagnosis and treatment of diseases, and therefore may have important applications in precision medicine.",
author = "Sevakula, {Rahul Kumar} and Vikas Singh and Verma, {Nishchal K.} and Chandan Kumar and Yan Cui",
year = "2018",
month = "4",
day = "3",
doi = "10.1109/TCBB.2018.2822803",
language = "English (US)",
journal = "IEEE/ACM Transactions on Computational Biology and Bioinformatics",
issn = "1545-5963",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - JOUR

T1 - Transfer Learning for Molecular Cancer classification using Deep Neural Networks

AU - Sevakula, Rahul Kumar

AU - Singh, Vikas

AU - Verma, Nishchal K.

AU - Kumar, Chandan

AU - Cui, Yan

PY - 2018/4/3

Y1 - 2018/4/3

N2 - The emergence of deep learning has impacted numerous machine learning based applications and research. The reason for its success lies in two main advantages: 1) it provides the ability to learn very complex non-linear relationships between features and 2) it allows one to leverage information from unlabeled data that does not belong to the problem being handled. This paper presents a transfer learning procedure for cancer classification, which uses feature selection and normalization techniques in conjunction with stacked sparse auto-encoders on gene expression data. While classifying any two tumor types, data of other tumor types were used in unsupervised manner to improve the feature representation. The performance of our algorithm was tested on 36 two-class benchmark datasets from the GEMLeR repository. On performing statistical tests, it is clearly ascertained that our algorithm statistically outperforms several generally used cancer classification approaches. The deep learning based molecular disease classification can be used to guide decisions made on the diagnosis and treatment of diseases, and therefore may have important applications in precision medicine.

AB - The emergence of deep learning has impacted numerous machine learning based applications and research. The reason for its success lies in two main advantages: 1) it provides the ability to learn very complex non-linear relationships between features and 2) it allows one to leverage information from unlabeled data that does not belong to the problem being handled. This paper presents a transfer learning procedure for cancer classification, which uses feature selection and normalization techniques in conjunction with stacked sparse auto-encoders on gene expression data. While classifying any two tumor types, data of other tumor types were used in unsupervised manner to improve the feature representation. The performance of our algorithm was tested on 36 two-class benchmark datasets from the GEMLeR repository. On performing statistical tests, it is clearly ascertained that our algorithm statistically outperforms several generally used cancer classification approaches. The deep learning based molecular disease classification can be used to guide decisions made on the diagnosis and treatment of diseases, and therefore may have important applications in precision medicine.

UR - http://www.scopus.com/inward/record.url?scp=85051168357&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85051168357&partnerID=8YFLogxK

U2 - 10.1109/TCBB.2018.2822803

DO - 10.1109/TCBB.2018.2822803

M3 - Article

JO - IEEE/ACM Transactions on Computational Biology and Bioinformatics

JF - IEEE/ACM Transactions on Computational Biology and Bioinformatics

SN - 1545-5963

ER -