Despite the increasing popularity, cognitive diagnosis models have been criticized for limited utility for small samples. In this study, the authors proposed to use Bayes modal (BM) estimation and monotonic constraint...
详细信息
Despite the increasing popularity, cognitive diagnosis models have been criticized for limited utility for small samples. In this study, the authors proposed to use Bayes modal (BM) estimation and monotonic constraints to stabilize item parameter estimation and facilitate person classification in small samples based on the generalized deterministic input noisy "and" gate (G-DINA) model. Both simulation study and real data analysis were used to assess the utility of the BM estimation and monotonic constraints. Results showed that in small samples, (a) the G-DINA model with BM estimation is more likely to converge successfully, (b) when prior distributions are specified reasonably, and monotonicity is not violated, the BM estimation with monotonicity tends to produce more stable item parameter estimates and more accurate person classification, and (c) the G-DINA model using the BM estimation with monotonicity is less likely to overfit the data and shows higher predictive power.
In this paper, we are concerned with a robustifying high-dimensional (RHD) structured estimation in finite mixture of multinomial models. This method has been used in many applications that often involve outliers and ...
详细信息
In this paper, we are concerned with a robustifying high-dimensional (RHD) structured estimation in finite mixture of multinomial models. This method has been used in many applications that often involve outliers and data corruption. Thus, we introduce a class of the multinomial logistic mixture models for dependent variables having two or more discrete categorical levels. Through the optimization with the expectation maximization (em) algorithm, we study two distinct ways to overcome sparsity in finite mixture of the multinomial logistic model;i.e., in the parameter space, or in the output space. It is shown that the new method is consistent for RHD structured estimation. Finally, we will implement the proposed method on real data. (C) 2021 The Authors. Published by Atlantis Press B.V.
Mean regression model could be inadequate if the probability distribution of the observed responses is not symmetric. Under such situation, the quantile regression turns to be a more robust alternative for accommodati...
详细信息
Mean regression model could be inadequate if the probability distribution of the observed responses is not symmetric. Under such situation, the quantile regression turns to be a more robust alternative for accommodating outliers and misspecification of the error distribution, since it characterizes the entire conditional distribution of the outcome variable. This paper proposes a robust logistic quantile regression model by using a logit link function along the em-based algorithm for maximum likelihood estimation of the pth quantile regression parameters in Galarza (Stat 6, 1, 2017). The aforementioned quantile regression (QR) model is built on a generalized class of skewed distributions which consists of skewed versions of normal, Student's t, Laplace, contaminated normal, slash, among other heavy-tailed distributions. We evaluate the performance of our proposal to accommodate bounded responses by investigating a synthetic dataset where we consider a full model including categorical and continuous covariates as well as several of its sub-models. For the full model, we compare our proposal with a non-parametric alternative from the so-called quantreg R package. The algorithm is implemented in the R package lqr, providing full estimation and inference for the parameters, automatic selection of best model, as well as simulation of envelope plots which are useful for assessing the goodness-of-fit.
In this paper, we discuss practical limitations of the standard choice-based demand models used in the literature to estimate demand from sales transaction data. We present modifications and extensions of the models a...
详细信息
In this paper, we discuss practical limitations of the standard choice-based demand models used in the literature to estimate demand from sales transaction data. We present modifications and extensions of the models and discuss data preprocessing and solution techniques which are useful for practitioners dealing with sales transaction data. Among these, we present an algorithm to split sales transaction data observed under partial availability, we extend a popular Expectation Maximization (em) algorithm for non-homogeneous product sets, and we develop two iterative optimization algorithms which can handle much of the extensions discussed in the paper.
Finite mixture models are one of the most widely used probabilistic methods for image segmentation. In this paper, we propose and investigate a mixture model based on Beta-Liouville distributions, which offers more fl...
详细信息
ISBN:
(纸本)9781728157306
Finite mixture models are one of the most widely used probabilistic methods for image segmentation. In this paper, we propose and investigate a mixture model based on Beta-Liouville distributions, which offers more flexibility than previously proposed models. The proposed approach is based on integration of mixture models with Markov Random Field (MRF) with a novel factor that is induced to reduce noise and illumination in images. The model is learned using Expectation Maximization (em) algorithm based on Newton-Raphson approach. The proposed approach is compared with mixtures of Gaussian, Dirichlet and generalized Dirichlet distributions with integrated MRF. The experimental results demonstrate that proposed segmentation framework gives better performance and better results as compared to mixtures of Gaussian, Dirichlet and generalized Dirichlet with MRF.
The problem of parameter estimation for nonlinear state-space models is addressed using the expectation-maximisation algorithm. Model states and parameters are iteratively estimated using cubature Kalman smoothing and...
详细信息
The problem of parameter estimation for nonlinear state-space models is addressed using the expectation-maximisation algorithm. Model states and parameters are iteratively estimated using cubature Kalman smoothing and maximum a posteriori estimation. A modification to this technique is proposed by weighting measurement samples so the algorithm equally tries to approximate all system dynamics, even those poorly represented in the measurements. The method is applied to parameter estimation of a vehicle dynamics model. Copyright (C) 2021 The Authors.
Tunnel fans are typical and key fire-fighting electromechanical equipment to ensure the ventilation and safety of tunnel traffic. Effective maintenance of such a group of complex electromechanical equipment servicing ...
详细信息
ISBN:
(纸本)9781665400497
Tunnel fans are typical and key fire-fighting electromechanical equipment to ensure the ventilation and safety of tunnel traffic. Effective maintenance of such a group of complex electromechanical equipment servicing in hazard environment is challenging for vulnerable to unexpected failure. However, the widely applied deep learning methods lack the capability that extracting features from sample organization. A novel semi-supervised graph neural network (ASGNN) is proposed that is adaptive to fluctuate fault features. First, a clustering method is proposed to develop a knowledge alignment layer for the construction of the graph. Then, the embedded representation of the graph network is introduced to aggregate the information of the whole graph. Finally, an expectation-maximization (em) algorithm-based learning method is developed to realize the alternate learning of both signal and relationship features. The proposed fault diagnosis solution has been verified with experiments, and the results demonstrated that the proposed method outperformed the state-of-the-art solutions.
The objective of Smart Manufacturing is to improve productivity and competitiveness in industry, based on in-process data. It requires reliable, explainable and understandable models such as Bayesian networks for perf...
详细信息
ISBN:
(数字)9783030794637
ISBN:
(纸本)9783030794620;9783030794637
The objective of Smart Manufacturing is to improve productivity and competitiveness in industry, based on in-process data. It requires reliable, explainable and understandable models such as Bayesian networks for performing tasks like condition prediction. In this context, a Bayesian network can be classically learned in a supervised, unsupervised way or a semi-supervised way. Here, we are interested in how to perform the learning when the ground truth isn't included in the learning data but is observable indirectly in another related dataset. This paper introduces a fully unsupervised variation of co-training that allows to include this second dataset, with two learning strategies (split and recursive). In our experiments, we propose one simple probabilistic graphical model used for predicting the state of a machine tool from results given by several sensors, and illustrate our unsupervised cotraining strategies first with benchmarks available from the UCI repository, for which 4 out of 5 datasets have best results with the recursive strategy. Finally, the recursive strategy was validated by McNemar's test as being the best strategy on a real industrial dataset.
In practical applications, non-Gaussianity of the signal at the sensor array is detrimental to the performance of conventional Direction-of-Arrival (DOA) estimators developed under the Gaussian model. In this paper, w...
详细信息
ISBN:
(纸本)9781728176055
In practical applications, non-Gaussianity of the signal at the sensor array is detrimental to the performance of conventional Direction-of-Arrival (DOA) estimators developed under the Gaussian model. In this paper, we propose a novel robust DOA estimator from the data collected at the sensor array under the corruption of non-Gaussian interference and noise. Additionally, the Cramer-Rao bound for DOA parameters under the considered signal model is derived. Simulation results show that the proposed estimator exhibits near-optimal estimation performance under the assumed model while being robust to model mismatch and/or the presence of outliers.
In this paper, we present a multichannel noise reduction scheme which uses the minimum variance distortionless response (MVDR) beamformer based on the second-order statistics (SOS) of the source and noise signals esti...
详细信息
ISBN:
(纸本)9781665432870
In this paper, we present a multichannel noise reduction scheme which uses the minimum variance distortionless response (MVDR) beamformer based on the second-order statistics (SOS) of the source and noise signals estimated in a frame-wise fashion. The time-frequency masks, required for SOS estimation, correspond to the probabilities of speech presence and speech absence in the noisy observations, and they are found by the proposed frame-based maximum a posteriori (MAP) estimator with an inverse Wishart prior distribution. The derived expectation-maximization (em) algorithm estimates the parameters of the assumed complex Gaussian mixture model (CGMM). The proposed approach is compared with an existing block-based method. Following the outline of mathematical differences between both processing schemes, we perform experimental evaluation. The obtained results indicate that the proposed frame-based approach outperforms block-based method by enabling stronger reduction of undesired noise, which in turn leads to better quality of the enhanced speech signal.
暂无评论