Literature Review

You are to look at the papers and links below as well as find 2 other papers on the topic of "Uncertainty Visualiztion". You are to read them and post to blackboar discussion group "Project 4- visualization research".

Understand your Goals

You should try to understand what the goals are of your system. I have outlined them below. Note that 2 and/or 3 are where you will be contributing new ideas beyond what you have seen in previous research

1) Using nearest neighbor recognize cancer/no cancer for the blob&feature processed image.

2) To clearly visualize the blob feature information.

3) Using potentially a number of techniques visualize the uncertanties presented.

4) Design a user-interface that lets the user better analyze the data and especially the uncertainty information via manipulation of how the uncertainty information and also possibly the input data is visualized/displayed

System Design

There are 3 main parts of your system design:

Recognition

Implement nearest neighbor utilizing training data YOU create with your system from Project 3 with the additional human entered in label (Cancer or No Cancer). Format of this information should be the same as your file for Project 3 with the addition for each blob the Cancer/No Cancer label.

Data Visualization

For each possible kind/format of blob feature data you will have to create a scheme to visualize it. This may be simple if the data is spatial. Pseudo-coloring can be a simple technique for non-visual non-spatial data.

Certainty + Importance Visualization

Figure out how you wish show on top of visualizing the feature itself its certainty and importance values.

Interface Design

Here you must decide what controls the user will have in viewing and manipulating the display of the visualiztion. See the project section "System Interface Functionality"

Implementation and Testing

This stage is what it says...simply the implementation, modification and testing of your system.

Delivery and Presentation

The deliverables are discussed on the Project's main page.