Algorithms
Dynamic Quantum Clustering, Marvin Weinstein
Computational Mathematics on Campus: ICME, Margot Gerritsen
The EMC algorithm: 3D reconstruction from noisy, unoriented, 2D single-particle diffraction data, Duane Loh, PULSE Institute, SLAC National Laboratory
A form of single-particle imaging has been experimentally realized where near-identical copies of a single target particle are interrogated (and destroyed) by diffractive imaging. Although each target particle produces an extremely noisy diffraction data of unknown orientation, a successful assembly of very many such data into the target's original 3D diffraction intensities not only reduces noise but can also in turn be used for structure determination. Using a simpler toy model, I will plainly describe how the EMC reconstruction framework [1] can accommodate such noise and orientational uncertainty: by requiring that all the collected 2D diffraction data fit agreeably with a set of common, reconstructed 3D intensities distribution. The ultimate challenge, naturally, is to recover such a set of 3D intensities distribution, with only the most general sample-independent assumptions. To conclude, I will show how this EMC framework was modified and already successfully applied to experimental diffraction data [2].
[1] N. D. Loh and V. Elser; Phys. Rev. E 80:026705, 2009.
[2] N. D. Loh, M. Bogan, V. Elser, A. Barty, S. Boutet, S. Bajt, J. Hajdu, T. Ekeberg, F. R. N. C. Maia, J. Schulz, M. M. Seibert, B. Iwan, N. Timneanu, S. Marchesini, I. Schlichting, R. L. Shoeman, L. Lomb, M. Frank, M. Liang, H. N. Chapman; Phys. Rev. Lett. 104, 225501, 2010.
Application of robotics algorithms for interpretation and model dynamics of protein structures, Liangjun Zhang
TMine and Classification Trees applied to Fermi Data, Alex Drlica-Wagner
Application of Classification Trees to Fermi Unidentified Sources, Maria Elena Monzani