Find the relationship between multiple independent variables to a dependent variable (GDP Growth %).
GDP = Consumption + Investment + Government Spending + Net Exports
Metrics: Adjusted R-Square is used to explain the degree to which predictor variables explain the variation of dependent variable while penalizing an increase of independent varibles
Assumptions: Variance Inflation Factor, Breusch–Pagan, Ljung-Box, Anderson-Darling Test
├── src
│ ├── MR_main.py # Multi Regression Model and loaded to parameters to pickle file
│ ├── MR_assumptions.py # Variance Inflation Factor, Breusch–Pagan, Ljung-Box, Anderson-Darling Test
│ ├── MR_data.py # Cleaned xlsx file and covert to pandas DataFrame
│ ├── MR_plot.py # Plot Variance Inflation Factor, Breusch–Pagan, Ljung-Box, Anderson-Darling Test Assumptionss
│ ├── MR_config.py # Define path as global variable
│ ├── LR_main.py # Initiating Linear Regression Model
│ ├── LR_metrics.py # Calculating metrics (R-Squared, MSE, MAE)
│ ├── LR_data.py # Extracted Adj-Closing price from YFinance
│ └── LR_config.py # Define path as global variable
├── inputs
│ ├── clean_data.csv # Cleaned file
│ ├── clean_data.csv # Cleaned file for Linear Regression Assets
│ └── korea_data.xlsx # Korea Economic data
├── plot
│ ├── Autocorrelation.png # Autocorrelation on data
│ ├── BoxPlot.png # Boxplot Features
│ └── ResidualMean.png # Residual Mean
├── requierments.txt # Packages used for project
└── README.md
OLS Regression Results
==============================================================================
Dep. Variable: GDP_growth R-squared: 0.892
Model: OLS Adj. R-squared: 0.880
Method: Least Squares F-statistic: 71.08
Date: Thu, 10 Jun 2021 Prob (F-statistic): 1.13e-19
Time: 11:48:40 Log-Likelihood: -84.898
No. Observations: 49 AIC: 181.8
Df Residuals: 43 BIC: 193.1
Df Model: 5 Features collected from WORLD BANK
Target
- GDP_growth (Annual %): Rate compares the year-over-year change in a country's economic output
Features
- Pop_growth (Annual %): Increase in the number of individuals in a population
- Broad_money_growth (Annual %): Measures economy's money supply (cash and other assets easily liquidated)
- Gov_consumtion_growth (Annual % growth): Aggregate transaction on a national income representing government expenditure on goods&services
- Gross_capital_formation_growth (% of GDP): Measured by the total value of the gross fixed capital formation
- Hh_consumption_growth (Annual % Growth): Value of all goods&services, purchased by households.
Linear Regression Model is the relationship between an independent variable (MSFT) to a dependent variable (SPY). Calculate the beta coefficient to measure the volatility of an individual stock compared to the systematic risk of the entire market.
Metrics: R-Square measure the proportion of the variance for a dependent variable that's explained by an independent variable in a regression model
Metrics: Mean Sqaured Error and Mean Absolute Error
Assumptions:
Slope: 1.47 and Intercept: 78.83
MSE 9.14e+02
MAE 25.8
RMSE 30.2
R2: 0.87Target
- S&P500 Index Fund: Index of 500 of the largest companies listed on US stock exchanges (Adjusted-Closing Price)
Feature
- Microsoft (MSFT): Adjusted-Closing Price