prior probability distribution 中文意思是什麼
prior probability distribution
解釋
先驗概率分佈- prior : n 普賴爾〈姓氏〉。adj 1 〈用作前置定[修飾]語〉在前的；優先的。2 〈與 to 連用〉在…之前 （opp post ...
- probability : n 1 或有；或然性。2 【哲學】蓋然性〈在 certainly 和 doubt 或 posibility 之間〉。3 【數學】幾率，...
- distribution : n 1 分配，分發，配給；分配裝置[系統]；配給品；配給量；【經濟學】配給方法，配給過程；分紅；【法律...
-
Compared with the regular rule - based expert system, the bayesian network based es can reason on the incomplete input information using the prior probability distribution ; the topological structure of the network being used to express the qualitative knowledge and the probability distributions of the nodes in the network being used to express the uncertainty of the knowledge, which made the knowledge representation more intuitively and more clearly ; applying the principle of the bayesian chaining rule, bidirectional inference which allow infer from the cause to the effect and from the effect to the cause can be achieved
與一般基於規則的專家系統相比，貝葉斯網專家系統利用先驗概率分佈，可以使推理在輸入數據不完備的基礎上進行；以網路的拓撲結構表達定性知識，以網路節點的概率分佈表達知識的不確定性，從而使不確定性知識的表達直觀、明確；利用貝葉斯法則的基本原理，可以實現由因到果及由果到因的雙向推理。 -
In addition, for general erlang ( n ) risk model, an integro - diifcrontial equation for the probability of ultimate ruin are presented : dickson arid hipp ( 2001 ) consider the erlang ( 2 ) risk model, and introduce the expectation of the discounted penalty h ' ( u ) which determines the joint and the marginal distribution of the time to ruin ( t ), the surplus prior to ruin ( u ( t - ) } and the deficit at ruin ( | u ( t ) | )
Dicksonandhipp （ 2001 ）同樣考慮了erlang （ 2 ）這種風險模型，並介紹了破產時的罰金折現期望w （ u ）這一概念。由罰金折現期望可得到破產時刻（ t ） ，破產前的瞬間盈餘（ u （ t ? ） ）和破產時的赤字（ u （ t ） ）的分佈和它們的聯合分佈，並給出了罰金折現期望滿足的一積分-微分方程，由此方程得到了罰金折現期望的拉普拉斯變換。 -
Monte carlo is a method that approximately solves mathematic or physical problems by statistical sampling theory. when comes to bayesian classification, it firstly gets the conditional probability distribution of the unlabelled classes based on the known prior probability. then, it uses some kind of sampler to get the stochastic data that satisfy the distribution as noted just before one by one
蒙特卡羅是一種採用統計抽樣理論近似求解數學或物理問題的方法，它在用於解決貝葉斯分類時，首先根據已知的先驗概率獲得各個類標號未知類的條件概率分佈，然後利用某種抽樣器，分別得到滿足這些條件分佈的隨機數據，最後統計這些隨機數據，就可以得到各個類標號未知類的后驗概率分佈。 -
In the sense of mean squares, maximum likelihood estimator, best linear unbiased estimator, taest linear invariant estimator, and good linear estimator are contracted. fourth, proposed and researched the reliability analysis method under the zero - failure data and doof data. based on the part beta distribution as the prior distribution of failure probability p, = p ( t < r, }, hierarchical bayesian estimate method was discussed, obtain the reliability analysis method under the zero - failure data and the doof data
第四，提出並研究了無失效數據類型和doof數據類型下電連接器的可靠性分析方法，提出了以不完全beta分佈為一級先驗分佈，超參數為[ 0 ， 1 ]上的均勻分佈作為失效概率先驗分佈的多層bayes方法，結合加權最小二乘法解決了產品在無失效數據和doof數據下的可靠性分析問題。 -
Combined with the prior distribution of the model parameters and water quality observation data, joint posterior probability function which stands for the distribution characters was obtained by bayes ' theorem
結合模型參數的先驗分佈和水質監測數據，通過貝葉斯定理計算獲得了表徵參數分佈規律的聯合后驗概率密度函數。 -
Finally, a new kind of methods on how to classify a sample into one of the several known populations in terms of posterior probability ratio established by the sample ' s predictive density functions when the unknown parameters " prior distributions are diffuse prior and minnesota prior or normal - inverted wishart distribution
最後，利用參數的充分統計量，根據后驗概率比構造了一類新的基於擴散先驗分佈和正態?逆wishart先驗分佈的多總體貝葉斯分類識別方法。 -
Secondly, revise factor coefficient with probability distribution, which given by experienced experts. thirdly, use bayes statistic deducing method to bind together the income rate of prior distribution and sample in formation, which makes forecast stocks in shenzhen stock market as samples. work out the series of weakly income rate
（ 2 ）對多元回歸的因子模型的各因子權重重做修正，將一些對金融市場有較透徹了解和豐富經驗的專家提供的信息引入，作出因子系數的概率分佈（並非隨意的主觀臆造） ，對模型的結果加以修正，以便提高模型的準確度。
分享友人