Daniel L. Elliott
Email: firstname.lastname@example.org, id:danelliottster
Twitter: danelliottrsch (no spaces)
Hard at work...
I am a Ph.D. student working under the tutelage of Professor Chuck Anderson. I really enjoy applying machine learning to actual, real world problems. I enjoy reinforcement learning, high dimensional data issues, mixture models, neural networks, and simple, yet effective, computer vision systems.
Some of my qualifications can be viewed in my resume.
My research, interests, and current work:
(please feel free to request code and data used in any of these experiments)
- Improving/speeding-up reinforcement learning through ancillary information:
- C.W. Anderson, Minwoo Lee and D.L. Elliott "Faster Reinforcement Learning After Pretraining Deep Networks to Predict State Dynamics," Proceedings of IJCNN 2015, July 12, 2015. (pdf) WINNER OF BEST PAPER AWARD
- D.L. Elliott and C.W. Anderson "Using supervised training signals of observable state dynamics to speed-up and improve reinforcement learning," Proceedings of ADPRL 2014, December 9, 2014. (pdf)
- Covariance regularization:
- Interests: random projections and classification, MPPCA
- D.L. Elliott, C. Anderson, M. Kirby "Covariance Regularization for Supervised Learning in High Dimensions," Proceedings of ANNIE 2010, November 1, 2010. (pdf)
- D.L. Elliott (2009). "Covariance Regularization In Mixture Of Gaussians For High-Dimensional Image Classification". MS Thesis, Colorado State University. (pdf)
- B. Draper, D.L. Elliott, J. Hayes, and K. Baek. " EM in High Dimensional Spaces ", IEEE Transactions on Systems, Man and Cybernetics, 35(3):571-577, June 2005. (pdf)
- Modelling grain drying and storage data:
- Current: Predicting seed drying times in forced-air seed drying buildings with heated air.
- D.L. Elliott, R.E. Valentine "Recurrent Neural Networks for moisture content prediction in seed corn dryer buildings," Proceedings of ICTAI 2011, November 7, 2011. (pdf|ps). Accompanying poster: (pdf|ps).
- ESN and Elman implementation based off code written by Elliott Forney.
Code and data sets
- Repository of my artificial neural network code can be found here.
- Image data used in many covariance regularization experiments: preprocessed to 32x32 and grayscale.
- The PyMix package. PyMix is a Python package for mixture models. I have contributed a little code to this excellent package. In particular, I added the ability to run the algorithm layed out in the ANNIE 2010 paper.
- Code I use with the SNOW package in R (much credit belongs to Andrew Sutton): snowfarm.R example.
Background obtained from: www.boogiejack.com
Last updated October 2014