線性辨別函數 的英文怎麼說

中文拼音 [xiànxìngbānbiéhánshǔ]
線性辨別函數 英文
linear discrimination function
  • 線 : 名詞1 （用絲、棉、金屬等製成的細長的東西） thread; string; wire 2 [數學] （一個點任意移動所構成的...
  • 性 : Ⅰ名詞1 （性格） nature; character; disposition 2 （性能; 性質） property; quality 3 （性別） sex ...
  • 辨 : Ⅰ動詞（辨別; 分辨; 明察） distinguish; discriminate; differentiate; recognize Ⅱ名詞（姓氏） a surname
  • 別 : 別動詞[方言] （改變） change （sb. 's opinion）
  • 函 : 名詞1. [書面語] （匣; 封套） case; envelope 2. （信件） letter 3. （姓氏） a surname
  • 數 : 數副詞（屢次） frequently; repeatedly
  • 線性 : [數學] [物理學] linear; linearity線性代數 linear algebra; 線性方程 linear equation; 線性規劃 line...
  • 辨別 : differentiate; distinguish; discriminate
  • 函數 : [數學] function函數計算機 function computer; 函數計算器 function calculator; 函數運算 functional operation
  1. Multi - channel deconvolution, based on multi - channel signal analysis, is presented to recover receiver function from three - component teleseismic p waveforms clustered within certain backazimuth and epicenter distance, to increase the resolution of receiver function. wavelet inversion is also introduced to broadband receiver function, to solve for nonuniqueness cased by linear approximation of nonlinear problem, and broadband receiver function is decomposed into multi - resolution versions, and the inversion result for the low - resolution version is iteratively taken as the initial model for the high - resolution, and gradually obtain the inversion model for broadband receiver function

    為克服廣義線性反演對初始模型的依賴性，並保持其快速收斂的特點，本文首次將小波變換引入到接收函數波形反演，將接收函數分解到不同的分辨尺度上，分別進行廣義線性反演，並將低解析度的反演結果作為高解析度反演的初始模型，漸近地獲取寬頻帶接收函數的模型，克服了非線性問題線性化帶來的非唯一性。
  2. This paper studies 3 kinds of algorithms : the viterbi algorithm, multiresolutional algorithm based on wavelet transformation and bayesian bootstrap algorithm. the viterbi algorithm is based on the hidden markov model theory and it is a kind of map estimation, this paper studies this algorithm and puts up an algorithm that suits for filtering in the presence of interference. multiresolutional algorithm takes full advantage of multiresolutional data, we can see it has a better filtering ability than the traditional filtering methods ; bootstrap algorithm is a recursive bayesian estimation, it describes the probability density function by the samples, so it can be used to nonlinear non - gaussion filtering, the simulation result of the two groundings is presented

    Viterbi演算法以隱馬爾可夫理論為基礎，是一種最大后驗概率估計方法，本文對該演算法進行了研究，給出了一種適合於非高斯干擾條件下的濾波方法；多分辨分析方法充分利用到了多解析度測量數據所包含的信息，從模擬結果中可以看出，該方法的濾波精度要高於傳統的濾波演算法；自主濾波方法是一種遞推貝葉斯估計演算法，它利用采樣點來描述目標狀態的概率密度函數，因而適用於非線性、非高斯條件下的濾波，本文分別對這兩種情況下的濾波進行了模擬。
分享友人
分享到 Line

分享到臉書