---
title: "etori (@enotori-in-japan) | Hashnode"
description: "Financial Engineering/Data Engineering/Data Scientist. Read the latest articles by etori on Hashnode."
url: "https://hashnode.com/@enotori-in-japan"
---

@enotori-in-japan

# etori

@enotori-in-japan · Japan · Joined February 2025

Financial Engineering/Data Engineering/Data Scientist

### [\[srm\] 主成分分析](https://enotori-in-japan-blog.hashnode.dev/srm-pca?utm_source=hashnode&utm_medium=feed)

主成分 樣本數n，第i個樣本 解釋變數p個，第j個變數 分解成p個主成分，第m個主成分，降維成M個 $$z\_{i,m}=\\phi\_{1,m}x\_{i,1}+\\phi\_{2,m}x\_{i,2}+...+\\phi\_{p,m}x\_{i,p}=\\sum\_{j=1}^{p}\\phi\_{j,m}x\_{i,j}$$$$\\begin{aligned} x\_{i,j}&=z\_{i,1}\\phi\_{1,j}+z\_{i,2}\\phi\_{2,j}+...+z\_{i,p}\\phi\_{p,j} \\\\&=\\sum\_{m=1}...

1 min read · [enotori-in-japan-blog.hashnode.dev](https://enotori-in-japan-blog.hashnode.dev/srm-pca?utm_source=hashnode&utm_medium=feed)

[Apr 11, 2025](https://hashnode.com/posts/srm-pca/67f893c472627f2073506576)

### [\[srm\] 決策樹](https://enotori-in-japan-blog.hashnode.dev/srm-decision-tree?utm_source=hashnode&utm_medium=feed)

單一決策樹 $$分割M-1次，得到M個節點R\_m，反應變數可為有K個水準的分類變數或連續變數$$節點不純度(Impurity) Classification error rate $$\\begin{aligned} E\_m &= \\frac{\\text{# misclassified}}{\\text{# samples in } R\_m} \\\\ &= 1 - \\max\_k(\\hat{p\_{mk}}) \\\\ \\text{(因為最大機率對應的 } &k \\text{ 就是 } \\hat{Y}\_m\\t...

1 min read · [enotori-in-japan-blog.hashnode.dev](https://enotori-in-japan-blog.hashnode.dev/srm-decision-tree?utm_source=hashnode&utm_medium=feed)

[Apr 11, 2025](https://hashnode.com/posts/srm-decision-tree/67f868e7a13fd0c9f4642f04)

### [\[Srm\] 時間數列](https://enotori-in-japan-blog.hashnode.dev/srm-time-series?utm_source=hashnode&utm_medium=feed)

弱平穩(Weak Stationary) $$\\begin{aligned} \\operatorname{E}(y\_t)&=C \\\\ \\operatorname{Cov}(y\_t,y\_s)&=R(t,s) \\\\ \\operatorname{Var}(y\_t)&=\\operatorname{Cov}(y\_t,y\_t)=\\sigma\_y^2 \\\\\\\\ \\rho\_k&=\\operatorname{Corr}(y\_t,y\_{t-k})=\\frac{\\operatorname{Cov}(y\_t,y\_{t-...

2 min read · [enotori-in-japan-blog.hashnode.dev](https://enotori-in-japan-blog.hashnode.dev/srm-time-series?utm_source=hashnode&utm_medium=feed)

[Apr 4, 2025](https://hashnode.com/posts/srm-time-series/67ef3b6820a919ee343882c6)

### [\[Srm\] 廣義線性模型](https://enotori-in-japan-blog.hashnode.dev/srm-generalized-linear-models?utm_source=hashnode&utm_medium=feed)

模型假設 Linear Exponential Family $$f(y;\\mu,\\theta)=\\operatorname{exp}(\\frac{y\\theta-b(\\theta)}{\\phi}-S(y,\\phi))$$$$\\begin{array}{c|c|c|c|c} & \\text{Bin}(n,\\pi) \\text{ (known } n\\text{)} & N(\\mu,\\sigma^2) & \\text{Poisson}(\\lambda) & \\text{Exp}(\\lambda) ...

3 min read · [enotori-in-japan-blog.hashnode.dev](https://enotori-in-japan-blog.hashnode.dev/srm-generalized-linear-models?utm_source=hashnode&utm_medium=feed)

[Mar 29, 2025](https://hashnode.com/posts/srm-generalized-linear-models/67e763c41a1deb3e1322b47d)

### [\[srm\] 線性迴歸](https://enotori-in-japan-blog.hashnode.dev/srm-linear-regression?utm_source=hashnode&utm_medium=feed)

符號定義 $$\\begin{aligned} {S\_{xx}} &:=\\sum(x\_i-\\bar{x})^2 \\\\ {S\_{xy}} &:=\\sum(x\_i-\\bar{x})(y\_i-\\bar{y}) \\\\ \\operatorname{SST} &=\\operatorname{TSS}=\\sum\\limits\_{i=1}^n (y\_i-\\bar{y})^2 \\\\ \\operatorname{SSR} &=\\operatorname{RegSS}=\\sum\\limits\_{i=1}^n (\\hat...

3 min read · [enotori-in-japan-blog.hashnode.dev](https://enotori-in-japan-blog.hashnode.dev/srm-linear-regression?utm_source=hashnode&utm_medium=feed)

[Mar 15, 2025](https://hashnode.com/posts/srm-linear-regression/67d5109389cd30ec583744ec)
