[數 Stat Machine Learning] 求教 2

係 又係我
其實係咪用Bayes Theorem 去prove
計到一半發現又無用佢個Hint 去做
完全迷失

條式一睇就知係 logistic regression
試下係依個方向prove

做到最後個exp(...)入面應該有啲terms係冇痴住支x
嗰啲terms夾埋就係bias term
如果define x_0 = 1嘅話就可以將個bias term收埋入支parameter vector θ入面
咁最尾exp(...)入面就會變咗純粹係兩支R^{N+1} vectors嘅dot product
好似題目入面寫咁樣

我想問吓樓主take緊乜course?
statistical machine learning?

係, machine learning

用乜textbook?
定係跟lecture notes?
我之前想幫吓你, 不過發現有啲terminology, algorithms唔識

跟lecture notes, 不過個prof 又唔教d 數點做 講完theory 就算

見到... 請教下你係咪都係用Bayes theorem 爆開normal distribution咁去計? 定係用其他計法

Bayes theorem硬爆

btw如果你唔知exp(...)入面邊啲term可以cancel嘅話
望一望LDA個wiki 裡面有寫