Interactive Visual Exploration and Analysis of High-Dimensional, Temporal, and Heterogeneous Biological Data
Abstract
High-dimensional data (hundreds of dimensions, or more) and temporal data (thousands of time frames) pose substantial challenges for both computational and interactive analysis. To reveal relevant intrinsic relations between items or dimensions, the utilization of only computational methods or standard visualization techniques is not enough. In this talk, we introduce the concept of interactive visual analysis (IVA) that enables us to combine computational methods with the user knowledge through a system of multiple linked views on the data and advanced interaction mechanisms. Our approach allows us to interact with the data on the level of individual items and also on the level of dimensions, exploiting a number of useful statistical methods in addition. To improve the understanding of temporal data, we utilize clustering methods, where the user is provided means to understand the internal cluster structure. Moreover, we also showcase how IVA can be beneficial when analyzing molecular dynamics.
J. Parulek, Interactive Visual Exploration and Analysis of High-Dimensional, Temporal, and Heterogeneous Biological Data, 2013.
[BibTeX]
High-dimensional data (hundreds of dimensions, or more) and temporal data (thousands of time frames) pose substantial challenges for both computational and interactive analysis. To reveal relevant intrinsic relations between items or dimensions, the utilization of only computational methods or standard visualization techniques is not enough. In this talk, we introduce the concept of interactive visual analysis (IVA) that enables us to combine computational methods with the user knowledge through a system of multiple linked views on the data and advanced interaction mechanisms. Our approach allows us to interact with the data on the level of individual items and also on the level of dimensions, exploiting a number of useful statistical methods in addition. To improve the understanding of temporal data, we utilize clustering methods, where the user is provided means to understand the internal cluster structure. Moreover, we also showcase how IVA can be beneficial when analyzing molecular dynamics.
@MISC {Parulek13Analysis,
author = "Julius Parulek",
title = "Interactive Visual Exploration and Analysis of High-Dimensional, Temporal, and Heterogeneous Biological Data",
howpublished = "Presentation in the VisBio 2013",
month = "September",
year = "2013",
abstract = "High-dimensional data (hundreds of dimensions, or more) and temporal data (thousands of time frames) pose substantial challenges for both computational and interactive analysis. To reveal relevant intrinsic relations between items or dimensions, the utilization of only computational methods or standard visualization techniques is not enough. In this talk, we introduce the concept of interactive visual analysis (IVA) that enables us to combine computational methods with the user knowledge through a system of multiple linked views on the data and advanced interaction mechanisms. Our approach allows us to interact with the data on the level of individual items and also on the level of dimensions, exploiting a number of useful statistical methods in addition. To improve the understanding of temporal data, we utilize clustering methods, where the user is provided means to understand the internal cluster structure. Moreover, we also showcase how IVA can be beneficial when analyzing molecular dynamics.",
images = "images/no_thumb.png",
thumbnails = "images/no_thumb.png",
location = "Bergen, Norway",
project = "physioillustration"
}