Finite mixture models have proven to be a powerful framework whenever unobserved heterogeneity cannot be ignored. We introduce in finance research the Mixture Hidden Markov Model (MHMM) that takes into account time and space heterogeneity simultaneously. This approach is flexible in the sense that it can deal with the specific features of financial time series data, such as asymmetry, kurtosis,...

1 Autoregressive models and Kalman filter models are Gaussian Markov and hidden Markov models, respectively 3 1.1 Example: voltage smoothing and interpolation; inferring biophysical parameters 5 1.2 We may perform inference in the Kalman model either via the forward-backwards recursion or by direct optimization methods . . . . . . . . . . . . . . . . . . . 9 1.3 The Kalman model is only identif...

The wavelet-domain Hidden Markov Tree Model can properly describe the dependence and correlation of fundus angiographic images' wavelet coefficients among scales. Based on the construction of the fundus angiographic images Hidden Markov Tree Models and Gaussian Mixture Models, this paper applied expectation-maximum algorithm to estimate the wavelet coefficients of original fundus angiographic i...

I use laboratory experiments to examine the relative performance of the English auction (EA) and the first price sealed bid auction (FPA) when procuring a commodity. The mean and variance of prices are lower in the FPA than in the EA. Bids and prices in EA agree with game theoretic predictions while they don’t in the FPA. To resolve these deviations found in the FPA, I consider a mixture model ...

We propose a low-memory-bandwidth, high-efficiency VLSI architecture for 60-k word real-time continuous speech recognition. Our architecture includes a cache architecture using the locality of speech recognition, beam pruning using a dynamic threshold, two-stage language model searching, a parallel Gaussian Mixture Model (GMM) architecture based on the mixture level and frame level, a parallel ...

In order to model the functional time series system, we developed a new model–Gaussian process hidden Markov model. We use the hidden Markov model to characterize the time order of system, and Gaussian process to model the function observations. We utilized this new model to consider the functional time series classification and prediction problem. The simulation results for real data demonstra...

In this paper we present an application of the read-once coupling from the past algorithm to problems in Bayesian inference for latent statistical models. We describe a method for perfect simulation from the posterior distribution of the unknown mixture weights in a mixture model. Our method is extended to a more general mixture problem, where unknown parameters exist for the mixture components...