ABSTRACT
Predictive diagnosis benefits both patients and hospitals. Major challenges limiting the effectiveness of machine learning based predictive diagnosis include the lack of efficient feature selection methods and the heterogeneity of measured patient data (e.g., vital signs). In this paper, we propose DLFS, an efficient feature selection scheme based on deep learning that is applicable for heterogeneous data. DLFS is unsupervised in nature and can learn compact representations from patient data automatically for efficient prediction. In this paper, the specific problem of predicting the patients' length of stay in the hospital is investigated in a predictive diagnosis framework which uses DLFS for feature selection. Real patient data from the pneumonia database of the National University Health System (NUHS) in Singapore are collected to verify the effectiveness of DLFS. By running experiments on real-world patient data and comparing with several other commonly used feature selection methods, we demonstrate the advantage of the proposed DLFS scheme.
- Anthony J. Bell and Terrence J. Sejnowski 1995. An Information-maximization Approach to Blind Separation and Blind Deconvolution. Neural Comput., Vol. 7, 6 (Nov. 1995), 1129--1159. binfopersonS. A. Mayer, J. Rivolta, J. Stillman, B. Boden-Albala, M. S. Elkind, R. Marshall, and J. Y. Chong. 2010. Impact of delayed transfer of critically ill stroke patients from the Emergency Department to the Neuro-ICU. Neurocrit Care, Vol. 13, 1 (Aug 2010), 75--81. Google ScholarDigital Library
- F. C. Ryckman, P. A. Yelton, A. M. Anneken, P. E. Kiessling, P. J. Schoettker, and U. R. Kotagal. 2009. Redesigning intensive care unit flow using variability management to improve access and safety. Jt Comm J Qual Patient Saf Vol. 35, 11 (Nov 2009), 535--543.Google Scholar
- Angela Simpson, Vincent Y. Tan, John Winn, Markus Svensen, Christopher Bishop, David Heckerman, Iain Buchan, and Adnan Custovic. 2010. Beyond Atopy: Multiple Patterns of Sensitization in Relation to Asthma in a Birth Cohort Study. American Journal of Respiratory and Critical Care Medicine Vol. 181 (February 2010), 1200--1206.Google ScholarCross Ref
- Ralph Snyderman. 2012. Personalized health care: From theory to practice. Biotechnology Journal, Vol. 7, 8 (2012), 973--979. 1860--7314Google ScholarCross Ref
- Pascal Vincent, Hugo Larochelle, Isabelle Lajoie, Yoshua Bengio, and Pierre-Antoine Manzagol. 2010. Stacked Denoising Autoencoders: Learning Useful Representations in a Deep Network with a Local Denoising Criterion. J. Mach. Learn. Res. Vol. 11 (Dec. 2010), 3371--3408. 1532--4435 Google ScholarDigital Library
- F. Wang, N. Lee, J. Hu, J. Sun, S. Ebadollahi, and A. F. Laine. 2013. A Framework for Mining Signatures from Event Sequences and Its Applications in Healthcare Data. IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 35, 2 (Feb 2013), 272--285. 0162--8828 Google ScholarDigital Library
- N. G. Weiskopf and C. Weng 2013. Methods and dimensions of electronic health record data quality assessment: enabling reuse for clinical research. J Am Med Inform Assoc Vol. 20, 1 (Jan 2013), 144--151.Google ScholarCross Ref
- Jionglin Wu, Jason Roy, and Walter F. Stewart 2010. Prediction Modeling Using EHR Data: Challenges, Strategies, and a Comparison of Machine Learning Approaches. Vol. 48, 6 (2010), S106--S113. 00257079Google Scholar
- F. Zhao and R. Yang. 2007. Power-Quality Disturbance Recognition Using S-Transform. IEEE Transactions on Power Delivery Vol. 22, 2 (April 2007), 944--950. 0885--8977Google ScholarCross Ref
- O. C. Zienkiewicz. 1976. Splines and variational methods. Internat. J. Numer. Methods Engrg. Vol. 10, 2 (1976), 487--487. 1097-0207Google ScholarCross Ref
Index Terms
- Learning Deep Representations from Heterogeneous Patient Data for Predictive Diagnosis
Recommendations
An agent-based game for the predictive diagnosis of parkinson's disease
AAMAS '14: Proceedings of the 2014 international conference on Autonomous agents and multi-agent systemsExisting Parkinson's Disease (PD) diagnosis relies heavily on doctors' observations combined with neurological exams. Such a technique is often inconvenient, infrequent, and subjective, which leads to a high misdiagnosis rate. As several cardinal ...
Incorporating Feature Selection Methods into Machine Learning-Based Covid-19 Diagnosis
AbstractThe aim of the study is to diagnose Covid-19 by machine learning algorithms using biochemical parameters. In addition to the aim of the study, October selection was performed using 14 different feature selection methods based on the biochemical ...
Kernel-based learning and feature selection analysis for cancer diagnosis
Graphical abstractDisplay Omitted HighlightsA novel feature selection approach is proposed based on two steps.First step uses SVM-RFE to prefiltre the gene; we select 60% of relevant genes.Second step uses Binary Dragon Fly algorithm to optimal subset ...
Comments