NOMAD is the alias for Non-locking, stOchastic Multi-machine framework for Asynchronous and Decentralized computation. This is a scalable distributed framework for various latent variable models. We have successfully applied NOMAD to two applications:
- Low-rank Matrix Factorization for Large-scale Recommender Systems
- Collapsed Gibbs Sampling for Large-scale Topic Modeling
- NOMAD for Low-rank Matrix Factorization [download]
- NOMAD for Collapsed Gibbs Sampling for LDA [download]
- A Scalable Asynchronous Distributed Algorithm for Topic Modeling (pdf, arXiv, software, code)
H. Yu, C. Hsieh, H. Yun, S. Vishwanathan, I. Dhillon.
In International World Wide Web Conference (WWW), pp. 1340-1350, May 2015. (Oral)
- NOMAD: Non-locking, stOchastic Multi-machine algorithm for Asynchronous and Decentralized matrix completion (pdf, software)
H. Yun, H. Yu, C. Hsieh, S. Vishwanathan, I. Dhillon.
In International Conference on Very Large Data Bases (VLDB), pp. 975-986, July 2014.