Link to the University of Pittsburgh Homepage
Link to the University Library System Homepage Link to the Contact Us Form

A computational method for the analysis of pain patterns and progression of pancreatitis with a large number of predictor variables

Tian, Ye (2014) A computational method for the analysis of pain patterns and progression of pancreatitis with a large number of predictor variables. Doctoral Dissertation, University of Pittsburgh. (Unpublished)

[img]
Preview
PDF
Submitted Version

Download (619kB) | Preview

Abstract

Chronic pancreatitis (CP) is a major burden of gastrointestinal disease in the United States accounts for significant healthcare costs to the society. Abdominal pain is the most common symptom in CP patients and the development of CP is challenging medical practice. It has been proposed that a combination of genetic, environmental, and metabolic risk factors contribute to the pain patterns in CP patients and development of CP. This research aimed to introduce a new data analytic strategy Random Forests (RF) to support big data analysis in studying CP and epidemiological researches.
RF has been becoming a popular non-parametric algorithm in computational method and used in many scientific areas in the context of big data era. RF is an ensemble of individual decision trees to help explore data structure and hidden information in high dimensional data. RF could deal with correlated predictor variables and integrates complex interaction effects during modeling process to evaluate the entire effects of all predictor variables on outcome variable and produce estimates of importance scores for all predictor variables.
In this work, a framework of combining RF analyses with traditional statistical analyses was developed to investigate important risk factors associated with different pain patterns in patients with CP and disease progression from recurrent acute pancreatitis (RAP) to CP. The public health significance of this novel analytic method is that it successfully examined a large amount of predictor variables in a multivariable way and would help researchers to better understand complex mechanisms in CP.


Share

Citation/Export:
Social Networking:
Share |

Details

Item Type: University of Pittsburgh ETD
Status: Unpublished
Creators/Authors:
CreatorsEmailPitt UsernameORCID
Tian, Yeyet3@pitt.eduYET3
ETD Committee:
TitleMemberEmail AddressPitt UsernameORCID
Committee ChairWisniewski, Stephen R.wisniew@edc.pitt.eduSTEVEWIS
Committee MemberBertolet, Marianne H.bertoletm@edc.pitt.eduMHB12
Committee MemberWhitcomb, David C.whitcomb@pitt.eduWHITCOMB
Committee MemberZmuda, Joseph M.zmudaj@edc.pitt.eduEPIDJMZ
Date: 27 June 2014
Date Type: Publication
Defense Date: 26 March 2014
Approval Date: 27 June 2014
Submission Date: 6 April 2014
Access Restriction: 5 year -- Restrict access to University of Pittsburgh for a period of 5 years.
Number of Pages: 116
Institution: University of Pittsburgh
Schools and Programs: Graduate School of Public Health > Epidemiology
Degree: PhD - Doctor of Philosophy
Thesis Type: Doctoral Dissertation
Refereed: Yes
Uncontrolled Keywords: pancreatitis, pain, random forests
Date Deposited: 27 Jun 2014 20:33
Last Modified: 01 May 2019 05:15
URI: http://d-scholarship.pitt.edu/id/eprint/21445

Metrics

Monthly Views for the past 3 years

Plum Analytics


Actions (login required)

View Item View Item