南京大学学报（自然科学版）
南京大學學報（自然科學版）
남경대학학보（자연과학판）
JOURNAL OF NANJING UNIVERSITY(NATURAL SCIENCES)
2009年
5期
689-698
,共10页
混合模型%变分学习%偏差信息准则%模型选择%鲁棒
混閤模型%變分學習%偏差信息準則%模型選擇%魯棒
혼합모형%변분학습%편차신식준칙%모형선택%로봉
mixture model%variational learning%deviance information criterion%model selection%robust
提出一种基于偏差信息准则(deriance information criterion,DIC)的鲁棒贝叶斯混合分布模型选择算法.在变分逼近框架下,给出鲁棒贝叶斯混合模型的DIC计算公式;设计的模型选择算法能同时学习模型参数推断和进行模型选择,避免在大的候选模型集中根据模型选择准则选取最优模型.给出试验参数初始值设置方法,在含有较多离群点的仿真数据和Old Faithful Geyser数据上的试验结果表明了好的性能:得到鲁棒的混合分量参数和较准确的混合分量个数.
提齣一種基于偏差信息準則(deriance information criterion,DIC)的魯棒貝葉斯混閤分佈模型選擇算法.在變分逼近框架下,給齣魯棒貝葉斯混閤模型的DIC計算公式;設計的模型選擇算法能同時學習模型參數推斷和進行模型選擇,避免在大的候選模型集中根據模型選擇準則選取最優模型.給齣試驗參數初始值設置方法,在含有較多離群點的倣真數據和Old Faithful Geyser數據上的試驗結果錶明瞭好的性能:得到魯棒的混閤分量參數和較準確的混閤分量箇數.
제출일충기우편차신식준칙(deriance information criterion,DIC)적로봉패협사혼합분포모형선택산법.재변분핍근광가하,급출로봉패협사혼합모형적DIC계산공식;설계적모형선택산법능동시학습모형삼수추단화진행모형선택,피면재대적후선모형집중근거모형선택준칙선취최우모형.급출시험삼수초시치설치방법,재함유교다리군점적방진수거화Old Faithful Geyser수거상적시험결과표명료호적성능:득도로봉적혼합분량삼수화교준학적혼합분량개수.
Bayesian approaches to robust mixture modelling based on Student-r distributions enable to be less sensitive to outliers, thereby preventing from over-estimating of the number of mixting components. However, there are two intractable problems in the previous methods for model selection under the variational Bayesian framework: (1) The variational approach converges to a local maximum of the low bound on the log-evidence that dependents on the initial parameter values. How can the variational approach guarantee that the initial settings for different models are consistency? (2) The low bound is sensitive to factorized approximation forms in the inference process. How can the variational approach guarantee that the approximate errors for different models are equivalent? In this paper, we present a model selection algorithm for robust bayesian mixture distributions based on deviance information criterion (DIC) proposed by Spiegelhalter et al. In 2002. Unlike the Bayesian Infromation Criterion (BIC), the DIC is straightforward in calculation, which has been adopted in many modern applications. Inspired by the works of McGrory et al. , which used the DIC values for model selection tasks of finite mixture Gaussian distributions and hidden Markov models, the calculation of a DIC for robust Bayesian mixture model is derived. The proposed algorithm can learn model parameters and perform model selection simultaneously, which avoids choosing an optimum one among a large set of candidate models. A method to initialize parameters of the algorithm is provided. Experimental results on simulated data and Old Faithful Geyser data containing a large amount of outliers show the good performance that the algorithm can learn parameters of mixture components robustly and the number of components precisely.
