Statistics Journal Club Winter Meeting

Friday, February 26, 2021 - 2:00 PM - 3:00 PM
Location
Contact

Michael Gray, mtgray@pdx.edu

Statistics Journal Club Winter Meeting

Dr. Naghmeh Daneshi will be giving a talk on "An Introduction to the Analysis of Clinical Trials". This talk will be very informative to anybody considering a career or doing research involving sensitive data from human subjects. 

We are meeting this Friday, February 26th at 2 pm via Zoom. The link to the Zoom meeting is https://pdx.zoom.us/j/81302629565

Here is the abstract of the talk:

There are certain protocols and trainings that need to be considered in any research involving human subjects. This talk will give a brief introduction about such research, human subjects and sensitive groups’ rights in the research. The talk is followed by introducing some famous designs in biostatistics. Another important piece in the data analysis of medical studies is preparation of a statistical analysis plan (SAP). I will introduce the basics and rationales of a SAP. Lastly, I will give an example to illustrate the difference between building a statistical model versus doing analysis using SAP.

Upcoming events

Allyson Meyers - STAT 501 Presentation

Allyson Meyers      will give a talk on the following statistical…
Add to my Calendar 2021-03-12 14:00:00 2021-03-12 15:00:00 Allyson Meyers - STAT 501 Presentation Allyson Meyers      will give a talk on the following statistical literature and problems project in partial fulfillment of requirements for the Master of Science in Statistics      “Comparative Analysis of Cerebrovascular Accident (CVA) Events using Survival Analysis and Bootstrap Methods” Abstract: This presentation will be based on an anonymous teaching dataset referencing the 1948 Framingham Heart Study, which was conducted by the National Heart, Lung, and Blood Institute. The main purpose of this presentation is to analyze right censored data using both survival analysis and bootstrap methods. The goal of this project is to study the survival time of Cerebrovascular Accident (CVA) events which are most commonly known as strokes. According to the American Heart Association (AHA), “a stroke occurs when the blood supply to the brain is blocked or when a blood vessel in the brain bursts (or ruptures), causing parts of the brain to become damaged or die” [16]. Characteristics of stroke patients, such as sex (SEX), age (AGE), use of cigarettes (SMOKER) and diabetes (DIAB) will all be studied and analyzed. The following survival analysis methods will be used to test the equality between two different sample groups: Kaplan-Meier estimates, Kaplan-Meier survival curves and the Log-Rank Test (LRT). In addition to this, hypothesis tests for the difference of means between each group will be tested and compared using a Two-Sample Z-test and Bootstrap-t Test. Each of these tests will then be analyzed using a standard normal critical value and bootstrap-t critical value at the 5% significance level. Finally, we will use a simulation of right-censored datasets (with known distributions and true means) to validate the accuracy and reliability of each hypothesis tested before. These simulations will be based on the overall censoring rate of the project dataset which was found to be 34%. The goal of this simulation is to determine the minimum sample size needed to obtain reliable and accurate results from a Z-test or Bootstrap-t test with right-censored data. Under the direction of Dr. Jong Sung Kim   Join Zoom Meeting https://pdx.zoom.us/j/9202656577 Kathie Leck Fariborz Maseeh Department of Mathematics + Statistics mathstatgrad@pdx.edu Kathie Leck Fariborz Maseeh Department of Mathematics + Statistics mathstatgrad@pdx.edu America/Los_Angeles public

Maggie Buffum - STAT 501 Presentation

Maggie Buffum      will give a talk on the following statistical literature and problems project in partial fulfillment of requirements for the…
Add to my Calendar 2021-03-12 13:00:00 2021-03-12 14:00:00 Maggie Buffum - STAT 501 Presentation Maggie Buffum      will give a talk on the following statistical literature and problems project in partial fulfillment of requirements for the Master of Science in Statistics  “Predicting Heating Systems from Hourly Electricity and Utility Customer Data” Abstract: The evaluation of the Utility's (kept anonymous) residential smart thermostat demand response (DR) programs for winter 2018/2019 revealed that the average electricity demand reduction per customer was less than expected and lower than those in winter 2017/2018. The evaluation suggested that changes in program eligibility screening for participants in fall 2018 resulted in the program enrolling residential customers who did not use electricity as their primary space heating fuel or enrolled their devices with non-controllable heating systems, which lowered the average demand reduction per customer and the program cost-effectiveness. To improve program performance and better target marketing efforts, the Utility seeks to find methods to identify candidate enrollees who do not use electricity as their primary space heating fuel. The Utility is also interested in distinguishing between heating system types, since not all electric heating systems are controllable under the current DR offerings or are expected to generate equal load impacts. The aim of this study is to develop a model that can accurately predict heating system types for individual Utility customers interested in participating in the DR program. For the study, the Utility provided pre-enrollment hourly consumption data and customer information for customers enrolled in the 2019/2020 program, all of whom enrolled under the 2018/2019 program eligibility screening methods but whose heating system types were later verified directly by the Utility. The analysis focuses on two predictive methods: multinomial logistic regression with hyperparameter tuning and random forest classifiers. Under the direction of Dr. Ge Zhao Join Zoom Meeting https://pdx.zoom.us/j/6268152440 Kathie Leck Fariborz Maseeh Department of Mathematics + Statistics mathstatgrad@pdx.edu Kathie Leck Fariborz Maseeh Department of Mathematics + Statistics mathstatgrad@pdx.edu America/Los_Angeles public