Research Buffet
Jeff Phillips
University of Utah
Monday, September 10, 2012
3147 MEB
Lecture 4:00 p.m.
Title: Accounting for Error in Large Data Sets
Jeff Phillips
University of Utah
Monday, September 10, 2012
3147 MEB
Lecture 4:00 p.m.
Title: Accounting for Error in Large Data Sets
Abstract
Due to the presence of enormous corpuses of data sets, the dominant scientific paradigm is changing from a hypothesis-driven collection of data to a data-mining-driven exploration of data. However, as data sets continue to grow, two fundamental challenges arise:
1. How do you summarize an enormous data corpus to a size manageable for deeper analysis?
2. How do you bound the error inherent in the data or introduced in the summarization phase?
I will provide fundamental techniques and analysis tools to deal with both of these questions, focusing in this talk on the broad class of data sets that can be interpreted as distributions. Specifically, I will show how to build a variety of distributions for statistics on uncertain data, how to analyze the approximations these techniques admit, and how to scale these techniques up to massive distributed data sets.
Tom Fletcher
University of Utah
Title: Image Analysis Track Overview & Manifold Statistics
Due to the presence of enormous corpuses of data sets, the dominant scientific paradigm is changing from a hypothesis-driven collection of data to a data-mining-driven exploration of data. However, as data sets continue to grow, two fundamental challenges arise:
1. How do you summarize an enormous data corpus to a size manageable for deeper analysis?
2. How do you bound the error inherent in the data or introduced in the summarization phase?
I will provide fundamental techniques and analysis tools to deal with both of these questions, focusing in this talk on the broad class of data sets that can be interpreted as distributions. Specifically, I will show how to build a variety of distributions for statistics on uncertain data, how to analyze the approximations these techniques admit, and how to scale these techniques up to massive distributed data sets.
Tom Fletcher
University of Utah
Title: Image Analysis Track Overview & Manifold Statistics
Abstract
First, I will give a brief overview of the Image Analysis Track:
Second, I will talk about my research in statistical analysis of manifold data. Manifold representations are useful for many different types of data, including directional data, transformation matrices, tensors, and shape. Statistical analysis of these data is an important problem in a wide range of image analysis and computer vision applications. However, defining statistics on a manifold is not a straightforward process. Even the simplest statistics, such as the mean, depend on the vector space structure of Euclidean space. This structure is not available for a general manifold. In this talk I will discuss how many common statistics can be defined for manifold-valued data by utilizing the geodesic distance on the manifold. After explaining the algorithms for computing statistics on manifolds, I will demonstrate their application in several image analysis problems.
First, I will give a brief overview of the Image Analysis Track:
Second, I will talk about my research in statistical analysis of manifold data. Manifold representations are useful for many different types of data, including directional data, transformation matrices, tensors, and shape. Statistical analysis of these data is an important problem in a wide range of image analysis and computer vision applications. However, defining statistics on a manifold is not a straightforward process. Even the simplest statistics, such as the mean, depend on the vector space structure of Euclidean space. This structure is not available for a general manifold. In this talk I will discuss how many common statistics can be defined for manifold-valued data by utilizing the geodesic distance on the manifold. After explaining the algorithms for computing statistics on manifolds, I will demonstrate their application in several image analysis problems.