KyungHyun Cho

M.Sc.,
Research Assistant in the Bayes Group
Research Assistant in the Machine Learning for Robotics Group
Department of Information and Computer Science,
Aalto University School of Science and Technology

Office:

Room A319 in Computer Science Building,
Konemiehentie 2, Otaniemi campus area, Espoo

Postal Address:

Aalto University School of Science,
Department of Information and Computer Science,
P.O. Box 15400, FI-00076 Aalto, Finland

Email:

firstname.lastname@aalto.fi

Code

1. Python (CUDA) Code for Restricted Boltzmann Machines: Download (updated: 08/02/2012)
   Implements Enhanced Gradient, Adaptive Learning Rate, and Parallel Tempering.
   CUV Library is required: Download
   liblinear Library is required for some functions: Download
   
2. MATLAB Code for Restricted Boltzmann Machines: Download
   Implements Enhanced Gradient, Adaptive Learning Rate, and Parallel Tempering.

Thesis

1. Cho, KyungHyun. Improved Learning Methods for Restricted Boltzmann Machines. Master's Thesis, Aalto University, 2011. [pdf]

Publication(s)

1. Cho, K. and Reyhani, N.
   An iterative algorithm for singular value decomposition on noisy incomplete matrices. [pdf]
   In Proceedings of the International Joint Conference on Neural Networks (IJCNN 2012), To appear, June 2012.
2. Cho, K., Raiko, T., and Ilin, A.
   Gaussian-Bernoulli Deep Boltzmann Machines. [pdf]
   NIPS 2011 Workshop on Deep Learning and Unsupervised Features Learning, Spain, Dec 2011.
3. Raiko, T., Cho, K., and Ilin, A.
   Derivations of the Enhanced Gradient for the Boltzmann Machine.
   Draft of a technical report, Aalto University, 2011.
   
4. Cho, K., Raiko, T., and Ilin, A.
   Enhanced Gradient and Adaptive Learning Rate for Training Restricted Boltzmann Machines. [pdf]
   In Proceedings of the International Conference on Machine Learning (ICML 2011), Bellevue, Washington, June 2011.
   Note: The enhanced gradients for visible and hidden biases that were mentioned to be approximate ones in this paper are the truly working enhanced gradients for them. Refer to the technical report above, this abstract, or my master's thesis to find a right definition of the enhanced gradients.
5. Cho, K., Ilin, A., and Raiko, T.
   Improved Learning of Gaussian-Bernoulli Restricted Boltzmann Machines. [pdf]
   In Proceedings of the International Conference on Artificial Neural Networks (ICANN 2011), Espoo, Finland, June 2011.
   Note: The enhanced gradient was used for all the experiments in this paper.
6. Raiko, T., Cho, K., Ilin, A.
   Enhanced Gradient for Learning Boltzmann Machines.
   In the Learning Workshop, Fort Lauderdale, Florida, April 2011. (abstract only)
7. Cho, K., Raiko, T., and Ilin, A.
   Parallel Tempering is Efficient for Learning Restricted Boltzmann Machines. [pdf]
   In Proceedings of the International Joint Conference on Neural Networks (IJCNN 2010), pp. 3246 - 3253, Barcelona, Spain, July 2010.