完整後設資料紀錄
DC 欄位語言
dc.contributor.author蔡嘉軒en_US
dc.contributor.authorTsai, Teresaen_US
dc.contributor.author洪慧念en_US
dc.contributor.authorHung, Hui-Nienen_US
dc.date.accessioned2014-12-12T02:33:22Z-
dc.date.available2014-12-12T02:33:22Z-
dc.date.issued2012en_US
dc.identifier.urihttp://140.113.39.130/cdrfb3/record/nctu/#GT070052603en_US
dc.identifier.urihttp://hdl.handle.net/11536/71776-
dc.description.abstractPolya 樹過程為一隨機測量之先驗分配,其中包含了Dirichlet 過程,亦即 Dirichlet 過程為Polya 樹過程的一個特例。此外,Polya 樹過程跟Dirichlet 過程一樣,都具有共軛分配的特質。但Polya 樹過程有兩點特徵與Dirichlet 過程不同,其一,Polya 樹過程具有大量參數共同決定分配,作為分配廣大的支撐;其二, Polya 樹過程具有近似連續分配的性質。zh_TW
dc.description.abstractPolya tree process is a large class of prior distributions, which Dirichlet process is included as a special case of Polya trees process. Like Dirichlet prior, the Polya tree prior also able to form a conjugate class for the posterior distributions. However, Polya tree process is different from Dirichlet process in two aspects: First, the Polya tree process is determined by a large collection of parameters and thus provide means to incorporate a wide range of beliefs. Second, some Polya trees assign probability 1 to the set of continuous distributions by (Kraft, 1964).en_US
dc.language.isoen_USen_US
dc.subjectPolya 樹先驗分配zh_TW
dc.subject貝氏無母數分析zh_TW
dc.subject隨機空間分割zh_TW
dc.subjectPolya tree prioren_US
dc.subjectBayesian nonparametricen_US
dc.subjectRandomised Partitionsen_US
dc.titlePolya 樹過程zh_TW
dc.titlePolya tree processen_US
dc.typeThesisen_US
dc.contributor.department統計學研究所zh_TW
顯示於類別:畢業論文