Categories

Archives

A new look at cartography

This website present various charts:

Radical Cartography

Share

Free Academic Books

An interesting website with open access academic books:

Free Academic Books

Share

Graphic Sociology

Analyzing the visual presentation of social data, an interesting blog:

Visualization of Social Data

Share

Predictive Analytics New York Conference: October, 2011

Click here to view the agenda at-a-glance.
Predictive Analytics World (PAW) October 16-21, 2011 in New York, USA, program is the richest and most diverse yet, featuring over 40 sessions across 3 tracks: 1) All Audiences, 2) Expert/Practitioner and 3) Financial Services

Share

Predictive Analytics London Conference: November, 2011

Click here to view the agenda at-a-glance.
Predictive Analytics World (PAW) 30 November – 1 December, 2011 in London, UK is packed with the top predictive analytics experts, practitioners, authors and business thought. PAW focuses on concrete examples of deployed predictive analytics. Hear from the horse’s mouth precisely how FT 500 analytics competitors and other top practitioners deploy predictive modelling, and what kind of business impact it delivers.

Share

Scientific Visualization

Bunkspeed Shot processed image of Spectramap Rhino Virus (OBJ file)

Bunkspeed Shot processed image of Spectramap Rhino Virus (OBJ file)

We have used Bunkspeed Shot to create a photo realistic image of a Spectramap Biplot from Rhinovirus test data. A Spectramap Biplot visualizes the contrast between row items (common-cold virus strains) and column items (antiviral compounds) of a data table. The data was published in: J Virol 1990 64 3 1117–1123 K Andries, B Dewindt, J Snoeks, L Wouters, H Moereels, P J Lewi, ‘Two groups of rhinoviruses revealed by a panel of antiviral compounds present sequence divergence and differential pathogenicity.’ The Spectramap Biplot can be thought of as a multidimensional scatter plot. Instead of three axes (XYZ) to scale three variables (e.g. in this example only three antiviral compounds), the biplot uses decomposed factors of both the columns and the rows of the original data table. Therefore, we are able to see all items in one plot, one 3D system of informational objects: cubes for columns and globes for rows. Note that they differ in size, which shows the mean potential of the antiviral compound or the mean sensitivity of the common-cold virus. Furthermore, color coding was applied to indicate the spectral contribution of the next three higher decomposed dimensions (i.e. factors 4, 5 and 6). Some of the items have their label displayed but all items have one. The Bunkspeed Shot photo realistic image adds a visually appealing additional sense of depth to the visualization. We hope soon to be able to have one published on the cover of a scientific journal!

Share

Spectramap CUDA Version

Currently, a Spectramap CUDA Version is under development for those users who need a high performance version. The focus is on the ability to analyse data tables with 100.000 cells or more. The first tests results are promising. On a Windows 7 system (ASUS GeForce 9600 GT Silent 512MB; Intel Core I7 2.8GHz 2.5 GT/s, 1/8MB) we have about 32fps with CUDA objects and about 100fps with OpenCL objects with 100.000 cells.

SPM CUDA with CUDA objects

SPM CUDA with CUDA objects

SPM CUDA with OpenCL Objects

SPM CUDA with OpenCL Objects

Share

Research subjects

Typically, visual analytics’ research subjects include query and reporting, decision support, analytical processing, statistical analysis, forecasting, and data mining. Visual analytics, like business intelligence, represents those systems that help companies understand what makes the corporation successful, and to help predict the future impact of current decisions. Applications are in the area of market analysis, customer profiling, customer contact analysis, market segmentation, credit scoring, product profitability, inventory movement, production metrics, statistics of sales, attendance and customer attrition, corporate governance, accounting & auditing, as well as certain risk related applications.

Share

Sense making and reasoning

Visual analytics is especially concerned with sense making and reasoning. It combines techniques from information visualization (that deals with abstract data structures) with techniques of data analysis, in particular transformation and decomposition, so that the research problem becomes manageable. Visual analytics is also an opportunity for the business community. It can facilitate evidence based decision making by means of extending the ‘box of tools’ of currently popular software, like spreadsheets. Having available data analysis tools that enable business analysis by visualizing hidden associations between variables could be beneficial for the decision making process. Moreover, the vast and ever growing amount of data stored in transactional and accounting information systems, drives the need for more (and better) analytical and representational tools.

Share

Visual analytics defined

Visual analytics is the science of analytical reasoning, an outgrowth of the field’s information visualization and scientific visualization, made possible by interactive visual software interfaces. Visual analytics’ tools and techniques are used to:
1) synthesize information and derive insight from dynamic, ambiguous, and often conflicting data;
2) detect the expected and discover the unexpected;
3) provide timely, defensible, and understandable assessments;
4) formulate research hypothesis as well as deliver evidence in their support; and
5) communicate assessments effectively for decision making and scientific progress.

Share