Displaying the Result and Recognizing Consulting Effort in Analyzing High-Dimensional Temporal and Spatial Correlated Data Set

• Activity Number: 260
• Type: Roundtables
• Date/Time: Tuesday, July 31, 2012 : 7:00 AM to 8:15 AM
• Sponsor: Section on Statistical Consulting
• Author(s): Hyun Kim*+
• Companies: University of California at Los Angeles
• Keywords: High-dimensional data ; Consulting ; Biomarker ; Correlated data ; Predictive ; Surrogate

Abstract

The need for analyzing complex statistical model due to high dimensional temporally and spatially correlated data set as a role of statistician/biostatisticians is ever increasing. Given these increased demands, recognition of these works can be summarized into two points: 1) a result needs to be interpretable to collaborators or medical investigators using graphical display like heat maps; 2) the result shall be consistent with the previously known outcome measurements. This will be an open discussion, facilitated by examining high-dimensional temporally and spatially correlated data set from CT and MRI images by the Center for Computer Vision and Imaging Biomarker, Department of Radiological Science and Department of Biostatistics at UCLA for researching a predictive classifier or early surrogate outcome as a role of a biomarker.