This is a demo visualization I built while teaching myself the D3 library
The data I'm plotting comes from an experiment I did where I applied machine learning to the problem of identifying heart disease from a series of physiological measurements. In that experiment I played with Perceptrons, Gaussian Mixtures, and Support Vector Classifiers, as well as various pre-processing techniques like PCA. The physiological data comes from a simple open dataset.
In total 3081 Models were trained, 25 times each for cross validation. Statistics on performance of each of these 3081 models across the 25 cross validation runs are displayed above.
My python code for the machine learning is available here
The javascript code for this visualization is available here
Data Available at the UCI machine learning repository here
Find me on LinkedIn here