| S. Chen, A. Kalai, A. Blum, and R. Rosenfeld. On-line algorithms for combining language models. 1998. In preparation. |
.... predicting disk idle times (Helmbold et al. 2000) and load balancing problems (Blum and Burch, 2000) as well as predicting TCP packet inter arrival times (Scott, 1998) Vovk (1997) developed a quasi probabilistic interpretation of the shifting expert algorithms by Herbster and Warmuth (1998a) Chen et al. 1998) and Singer (1998) signi cantly expanded the shifting expert algorithms to quasiprobabilistic methods for combining language models and managing portfolios, respectively. The rest of this article is outlined as follows. In Section 2, we give a general overview of how a certain class of on line ....
S. Chen, A. Kalai, A. Blum, and R. Rosenfeld. On-line algorithms for combining language models. Submitted to International Conference on Acoustics, Speech, and Signal Processing (ICASSP '99), 1998.
No context found.
S. Chen, A. Kalai, A. Blum, and R. Rosenfeld. On-line algorithms for combining language models. 1998. In preparation.
No context found.
A. Kalai, S. Chen, A. Blum, and R. Rosenfeld. On-line Algorithms for Combining Language Models. Proceedings of the International Conference on Accoustics, Speech, and Signal Processing, 2:745-748, 1999.
.... 5, 6] We present an efficient implementation of the Universal algorithm that is based on non uniform random walks that are rapidly mixing [1, 12, 7] This same implementation also works for non financial applications of the Universal algorithm, such as data compression [6] and language modeling [10]. 1. Introduction A constant rebalanced portfolio (CRP) is an investment strategy which keeps the same distribution of wealth among a set of stocks from day to day. That is, the proportion of total wealth in a given stock is the same at the beginning of each day. Recently there has been work on ....
....it is especially important to achieve this expectation. For example, a 50 chance at 10 million dollars may not be as valuable to most people as a guaranteed 5 million dollars. While our implementation can be used for applications of UNIVERSAL, such as data compression [6] and language modeling [10], we do not implement it in the case of transaction costs [2] or for the Dirichelet(1=2; 1=2) UNIVERSAL [4] ....
A. Kalai, S. Chen, A. Blum, and R. Rosenfeld. On-line Algorithms for Combining Language Models. In Proceedings of the International Conference on Accoustics, Speech, and Signal Processing (ICASSP '99), 1999.
....they are equally likely for any topic. This paper introduces the notion of nonlinearly interpolating the predictions from a general and a topicspecific language model to boost the probabilities of on topic words and suppress the probabilities of off topic words. Previous work in topic adaptation [1, 3, 4, 5, 7, 10, 11] has mainly focused on identifying topic specific subsets of the training text and building language models from them. The topic language models are linearly interpolated with a general language model built from all of the training text. Using this technique, all of the available models are ....
S. Chen, A. Kalai, A. Blum, and R. Rosenfeld. On-line algorithms for combining languagemodels. 1998. In preparation.
No context found.
, Phoenix, AZ, 1999.
No context found.
Kalai, A., Chen S.,, Blum A., and Rosenfeld R.: On-line Algorithms for Combining Language Models. Proceedings of the International Conference on Acoustics, Speech, and Signal Processing (ICASSP), 1999.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC