HappinessScoreRegression

Construct multivariate linear regression to predict the determinant factors for the Happiness Score in World's Happiness Report 2015

1. Data collection and preparation

• The data was collected from Kaggle.com based on World Happiness Report 2015
• 158 countries ranked by happiness scores following by 6 factors
• Remove all the NA values
• Investigate the predictors

2. Model selection

• AIC test (assume linear relation w/ interactions between factors)
• Linear Regression model based on correlation
  • ANOVA(full, reduced)

3. Model diagnosis

• Residual properties
• Residual normality
• Residual vs. predictors
• Constancy of error variance
• Data vs. model
• Outlier & Leverage

4. Model validation

• Used the fitted model to predict 2016 data and compare with the actual 2016 data

Sample graphical analysis

Comparison between predicted result and actual 2016 data