1. What were my goals for this week?
- Get moleculeLoader working with my project
- Write a routine to spit out the clusters into a PDB for viewing in PyMol or somesuch
- Test algorithms with position data to see if they work
- Look at possible ways to extend this to take motions into account?
2. What did I accomplish this week?
- Got moleculeLoader working with my project
- Got clusters into PDB format for VMD and PyMol
- Tested k-means with position data for alanine - it works
- Began reading about mean-shift algorithms, picked up the implementation FAMS
- Began compiling list of features of different clustering algorithm (dis)advantages for our purposes with motions
3. Why are numbers 1 and 2 different?
This week they're actually not that different at all! I'm proud of myself.
4. What are my goals for next week?
- Write a sanity checker for the molecule reader - right atoms/bonds/etc
- Extend Aaron's GNM to elastic and chemical network models
- Look at how to qualitatively analyze motions produced by those models
- Look at how to proceed more specifically on grouping for abstractions
5. How does this fit in my bigger picture?
The first gets my hands dirty in code - always a good thing. The others will help to make some decisions about the implementation of the motion abstraction project and hopefully help us to generate some of our own data without elNemo.
6. What did I read this week?
- Comaniciu, Dorin and Meer, Peter, Mean Shift: A Robust Approach Toward Feature Space Analysis.
- Fukunaga, Keinosuke and Hostetler, Larry D., The Estimation of the Gradient of a Density Function, with Applications in Pattern Recognition.
- Georgescu, Bogdan and Shimshoni, Ilan and Meer, Peter, Mean Shift Clustering in High Dimensions: A Texture Classification Example
::