PhD筆記03

W01 Introduction

2023-09-13三 14:10-17:00 林師模教授 shihmolin@gmail.com

Standard Deviation - Explained and Visualized

同學: TA陳mr.Jonal(會計), 孫娟Jen(國貿/福州)Shelro, Romio 羅汨; Reyner 倪悠思; and Assistant-Ms.陳Yachu(Line)
參考資料: Econometrics 5th
For reference
買到林師謨老師著多變量分析-管理上的應用拜讀。(二手書)

1-5章介紹方法、工具與數學運算。
6-10章討論了5種分析方法。

1. [主成份分析] (Principal Component Analysis, PCA): 這種分析方法通過將數據轉換到新的坐標系統中，來降低數據的維度，同時保留最重要的變異性信息。
2. [因素分析] (Factor Analysis): 這是一種統計方法，用於識別觀察變量背後的潛在關係，通常用於數據縮減和變量的分類。
3. [區別分析] (Discriminant Analysis): 此分析技術用於模型建立，以預測或區分兩個或多個自然或預先定義的組別。
4. [典型相關分析] (Canonical Correlation Analysis): 這種分析用於研究兩組變量之間的相互關係，找出一對變量集合，使得它們之間的相關性最大化。 | [舉例] |
5. [集群分析] (Cluster Analysis): 這是一種將對象分組的技術，使得組內對象之間非常相似，而組間對象則相異。

11章討論問卷調查
12章討論信度效度

P value Why the p-value is significant
The basic of Statistics 統計學基本-複習
mean算術平均數Arithmetic mean, 幾何平均數Geometric mean, 調和平均數Harmonic mean的算法與用法。
3M算術平均數mean、中位數median、眾數mode之比較有影片說明，有影片說明。
Variable 何謂變數

Course Introduction 課程介紹 :

It's all about Econometrics,計量經濟學, in this course.

Quantitative Methods I:
2 kinds methos: there are two category:
1.Econometrics, periodical (our focus, this semester 本學期重點)
2.Multivariate Analysis: Factor analysis, Principal component analysis, custer analysis. (little chance to mention this)

Schedule
w01 Introduction
w02 Review of Basic Statistics
w03 Analysis of Variance -> Hypothesis
w04 Simple Linear Regression Model(1) 請假absent
w05 Simple Linear Regression Model(2) 請假absent
w06 Interval Estimation and Hypothesis Testing(1)
w07 Interval Estimation and Hypothesis Testing(2)
w8 Goodness-of-fit and modelling Issues(1)
w09 Mid-tern report (Home work)
w10 Goodness-of-fit and modelling Issues(2)
w11 The Multiple Regression Model
w12 Furterinference in the Multiple Regression Model(1)
w13 Furterinference in the Multiple Regression Model(2)
w14 Using Indicator variables(1) dummy vareable
w15 Using Indicator variables(2)
w16 Heteroskedasticity(1) 異方差性
w17 Heteroskedasticity(2) 異方差性
w18 Final Report (Final Report summit)

Why the p-value is significant
If the p-value is 0.05 or lower, the result is trumpeted as significant, but if it is higher than 0.05, the result is non-significant and tends to be passed over in silence.

How do you write an inference for regression analysis?
ˆy=a+bx is an estimate for μy=α+βx. We conduct statistical inference in linear regression when we find a sample slope and then use it to make a confidence interval or perform a hypothesis test about the true population slope.
Heteroskedasticity 異方差性

Bing簡介Econemetrics
我從網路上搜尋到了一些關於Econometrics的資訊，並根據它們來簡介這個學科。
Econometrics是一門應用統計學和數學方法來分析經濟數據的學科。它的目的是為了給經濟關係提供實證內容，測試經濟理論，發展經濟模型，分析經濟歷史，和預測未來趨勢)。從事Econometrics的人被稱為Econometricians。

Econometrics的基本工具是多元線性回歸模型。它是一種用來描述一個或多個自變量（解釋變量）和一個因變量（被解釋變量）之間的關係的數學方程式。例如，我們可以用多元線性回歸模型來估計國民生產總值。例如，我們可以用多元線性回歸模型來估計國民生產總值（GDP）與失業率、利率、通貨膨脹率等因素之間的關係。
Econometrics的理論部分使用統計理論和數學統計來評估和發展Econometrics方法。Econometricians嘗試找到具有良好統計性質的估計量，例如無偏性、有效性和一致性。Econometrics的應用部分使用理論Econometrics和實際數據來評估經濟理論，發展Econometrics模型，分析經濟歷史，和預測未來趨勢。
EViews kind of statistics software.[EViews](https://zh.wikipedia.org/zh-tw/EViews) 專門為Econometics 專為統計的可以用SAS 或SPSS。
ANOVA 可以用Excel來計算 basic觀念。

Example of major issue of this class:
try to find out the Relationship between variables
50 students-have a test- every one got a grade of Math and English
if the teacher would like to know the relationship between Math and English
How to find out is it relationship, or not? if there is what is that?
Discrete or continued
positive slop = positive relationship

Confident is also a issue.

correlation coefficient rabge between: negative-1 <= x <= 1 positive: relationship

0,8 -> correlation
要說是不是significant 要檢定testing
relationship= perfect(certain) - strong - medium - weak
what kind of methods would be apropreate to fit your need.

想到一個問題是最難的就是想到一個idea.
collect all the data, run the regression to get number, check the relation.
do the test.
to testify economic theory.
this method is quantitative method

Forecasting: 2half for time serials data (cross section data)
什麼是time serials data 和 cross sectional data 有何異同

cross sectional data: collect manny many subject 'at the same time'
at 13:00 clossing clock, at that time all the price is the time serials data: the daily closing price of TSMC of past one year.
At this semester will concentrate to Cross sectional data. (net sem. will be SD).

What is Regression Analysis
- It is a statistical technique that attempts to 'explain' movements in one variable. the dependent variables, as a function of movements in a set of other variables. called the independent (or explanatory ) variable

- Economists are interested in cause-and effect.
- Don't be deceived by the words 'depentent' and 'independent' variables.
- Regression results cannot prove causelity! 回歸結果不能證明因果關係

- For example, if variables A and B are related statistically, the A might 'cause' B and B might 'cuase' A
some third factor might ca both might happen by chance

我的系統是ubunt22.04 請教我怎樣安裝統計軟體Eviews
使用方法為:(記住要先斷網無wifi連接)

W02 Review of Basic Statistic

2023-09-21三 14:10-17:00 林師模教授

For reference: Review of Statistics should study in advanced.
Reference:subject should study