Dimensionality reduction: feature extraction & feature selection
Technical Approach : Time-series preprocessing and feature extraction
-
Upload
jessamine-tillman -
Category
Documents
-
view
16 -
download
0
description
Transcript of Technical Approach : Time-series preprocessing and feature extraction
Project Title: Integrated Real-Time Clinical Deterioration Prediction for Hospitalized Patients and Outpatients
Technical Approach:•Time-series preprocessing and feature extraction•Nonlinear, interpretable, scalable, and sparse time-series classification algorithms;•Data-driven dynamic sensing reconfiguration schemes;
T. Bailey, Y. Chen, Y. Mao, C. Lu, G. Hackmann, S. T. Micek, K. Heard, K. Faulkner, and M. H. Kollef, “A Trial of a Real-time Alert for Clinical Deterioration in Patients Hospitalized on General Medical Wards.” Journal of Hospital Medicine, 8(5):236-242, 2013.
http://www.cse.wustl.edu/~ychen/public/hospital.pdf
Motivation:•To develop integrated early warning system alerts for hospitalized patients employing both electronically available data and real-time vital signs data;•To establish predictive algorithms from informational cloud data for outpatients potentially at risk for hospital readmission;Transformative:• New algorithms for predicting clinical deterioration and readmission from noisy and high-dimensional data streams;• A novel alert explanation system which highlights the most relevant factors and suggests possible intervention;• An adaptive sensing reconfiguration scheme for better user experience and energy efficiency;
Broader Impacts:• Improved clinical outcomes and potential reduction of both patient mortality rates and healthcare costs;• Widely-disseminated methods and software;• It addresses a major medical, societal, and governmental concern which has the potential to improve the overall administration of healthcare in the United States.