Inferential Analysis on Economic Metrics of the Nations of the World

As we can see, the data is presented to us in one big chunk. Looking at the attributes of the data set, it becomes apparent that we can divide this big table into smaller, logical, modular pieces. I have tried to partition the parent table logically into 3 tables.

In summary, our data set is now composed of 4 tables as follows:

1. Countries_Geography: This table stores information about the geographical attributes of countries.
2. Countries_Metrics: This table stores information about economic metrics of the countries.
3. Contries: This table stores general(miscellaneous) information about the Countries
4. Continent: This is the Continent-Country mapping table.

A glimpse of our tables-

Local <- odbcConnect("R Connect" ,uid = "", pwd = "")
Countries_Metrics <- sqlQuery(Local, "select * from Countries_Metrics")
Countries <- sqlQuery(Local, "select * from Countries")
Continents <- sqlQuery(Local, "select * from Continents")
Countries_Geography <- sqlQuery(Local, "select * from Countries_Geography")
kable(names(Countries))

kable(str(Countries))

## 'data.frame':    197 obs. of  5 variables:
##  $ Country_Code              : Factor w/ 196 levels "AD","AE","AF",..: 1 2 3 4 5 6 7 8 9 10 ...
##  $ Country_Name              : Factor w/ 197 levels "Afghanistan",..: 4 186 1 6 2 8 5 7 10 9 ...
##  $ Government_Type           : Factor w/ 52 levels "a parliamentary democracy, a federation, and a constitutional monarchy",..: 37 24 28 10 37 43 43 43 21 19 ...
##  $ Founding_Independence_Date: Date, format: "1278-01-01" "1971-12-02" ...
##  $ Capital_City              : Factor w/ 197 levels "Abu Dhabi","Abuja",..: 8 1 74 156 179 196 94 41 187 44 ...

kable(head(Countries))

kable(names(Countries_Geography))

kable(str(Countries_Geography))

## 'data.frame':    197 obs. of  6 variables:
##  $ Country_Code     : Factor w/ 196 levels "AD","AE","AF",..: 3 5 49 1 7 4 8 6 10 9 ...
##  $ Capital_Latitude : Factor w/ 194 levels "0 13 S","0 19 N",..: 96 126 104 131 187 44 97 122 100 149 ...
##  $ Capital_Longitude: Factor w/ 196 levels "0 05 W","0 13 W",..: 158 67 91 4 30 153 142 120 40 49 ...
##  $ Area             : num  652230 28748 2381741 468 1246700 ...
##  $ Land_Boundaries  : num  5987 691 6734 118 5369 ...
##  $ Coastline        : num  0 362 998 0 1600 ...

kable(head(Countries_Geography))

kable(names(Countries_Metrics))

kable(str(Countries_Metrics))

## 'data.frame':    197 obs. of  17 variables:
##  $ Country_Code                          : Factor w/ 196 levels "AD","AE","AF",..: 3 5 49 1 7 4 8 6 10 9 ...
##  $ Population                            : num  31822848 3020209 38813722 85458 19088106 ...
##  $ Life_Expectancy                       : num  50.5 78 76.4 82.7 55.3 ...
##  $ GDP                                   : num  4.53e+10 2.83e+10 2.85e+11 3.16e+09 1.32e+11 ...
##  $ Literacy_Rate                         : num  28.1 96.8 72.6 100 70.4 99 97.9 99.6 99 98 ...
##  $ Health_Expenditure_per_GDP            : num  9.6 6.3 3.9 7.2 3.5 5.9 8.1 4.3 9 10.6 ...
##  $ Military_Expenditure_per_GDP          : num  4.74 1.47 4.48 NA 3.63 NA 0.91 3.92 1.71 0.81 ...
##  $ Renewable_Water_Resources             : num  65.3 41.7 11.7 NA 148 ...
##  $ Net_Migration_Rate_per_1000_population: num  -1.83 -3.31 -0.93 0 0.47 2.23 0 -5.88 5.74 1.76 ...
##  $ Labor_Force                           : num  7512000 1098000 11150000 36060 9018000 ...
##  $ Internet_Users                        : num  1000000 1300000 4700000 67100 606700 ...
##  $ Commercial_Prime_Lending_Rate         : num  15 9.52 8 NA 15 10.3 16.4 16.5 6.2 2.2 ...
##  $ GDP_per_Capita                        : num  1424 9383 7335 37012 6905 ...
##  $ Economy_Status                        : Factor w/ 3 levels "Developed","Developing",..: 3 2 2 1 2 1 1 2 1 1 ...
##  $ Labor_force_participation_rate        : num  0.236 0.364 0.287 0.422 0.472 ...
##  $ Technology_Penetration_rate           : num  0.0314 0.4304 0.1211 0.7852 0.0318 ...
##  $ Population_Density                    : num  48.8 105.1 16.3 182.6 15.3 ...

kable(head(Countries_Metrics))

kable(names(Continents))

kable(str(Continents))

## 'data.frame':    197 obs. of  3 variables:
##  $ Country Name: Factor w/ 197 levels "Afghanistan",..: 3 5 19 23 27 28 30 32 33 34 ...
##  $ Continent   : Factor w/ 5 levels "Africa","America",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ Country_Code: Factor w/ 196 levels "AD","AE","AF",..: 49 7 20 26 16 19 36 41 31 169 ...

kable(head(Continents))

Analysis on Literacy rates across different Continents and Countries

Literacy_Rates_Continents <- sqlQuery(Local,"select continent, max(Literacy_Rate) as Highest_Literacy_Rate, min(Literacy_Rate) as Lowest_Literacy_Rate, Avg(Literacy_Rate) as Average_Literacy_Rate
from Countries_Metrics, Continents where Countries_Metrics.Country_Code=Continents.Country_Code
group by continent order by continent")

Highest_literacy_Rate_Country <- sqlQuery(Local, "select Country_Name from Countries,Countries_Metrics
where Literacy_Rate in (select max(Literacy_Rate) from Countries_Metrics) 
and Countries.Country_Code=Countries_Metrics.Country_Code")

Lowest_literacy_Rate_Country <- sqlQuery(Local, "select Country_Name from Countries,Countries_Metrics
where Literacy_Rate in (select min(Literacy_Rate) from Countries_Metrics) 
and Countries.Country_Code=Countries_Metrics.Country_Code")

Literacy_Rates_Max_Min <- sqlQuery(Local,"select max(Literacy_Rate) as Highest_Literacy_Rate, min(Literacy_Rate) as Lowest_Literacy_Rate, Avg(Literacy_Rate) as Average_Literacy_Rate
from Countries_Metrics")

Literacy_Highest_lowest <- list()
Literacy_Highest_lowest[[ "Highest_Country" ]] <- Highest_literacy_Rate_Country$Country_Name
Literacy_Highest_lowest[[ "Lowest_Country" ]] <- Lowest_literacy_Rate_Country$Country_Name
Literacy_Highest_lowest[["Highest_Value"]] <- Literacy_Rates_Max_Min$Highest_Literacy_Rate
Literacy_Highest_lowest[["Lowest_Value"]] <- Literacy_Rates_Max_Min$Lowest_Literacy_Rate
Literacy_Highest_lowest[["Average Value"]] <- Literacy_Rates_Max_Min$Average_Literacy_Rate

kable(Literacy_Rates_Continents)

Literacy_Highest_lowest

## $Highest_Country
## [1] Andorra       Finland       Greenland     Korea, North  Liechtenstein
## [6] Luxembourg    Norway       
## 7 Levels: Andorra Finland Greenland Korea, North ... Norway
## 
## $Lowest_Country
## [1] South Sudan
## Levels: South Sudan
## 
## $Highest_Value
## [1] 100
## 
## $Lowest_Value
## [1] 27
## 
## $`Average Value`
## [1] 84.85699

Analysis on Life Expectancy rates across different Continents and Countries

Life_Expectancy_Rates_Continents <- sqlQuery(Local,"select continent, max(Life_Expectancy) as Highest_Life_Expectancy, min(Life_Expectancy) as Lowest_Life_Expectancy, Avg(Life_Expectancy) as Average_Life_Expectancy
from Countries_Metrics, Continents where Countries_Metrics.Country_Code=Continents.Country_Code
                                             group by continent order by continent")

Highest_Life_Expectancy_Rate_Country <- sqlQuery(Local, "select Country_Name from Countries,Countries_Metrics
where Life_Expectancy in (select max(Life_Expectancy) from Countries_Metrics) 
                                                 and Countries.Country_Code=Countries_Metrics.Country_Code")

Lowest_Life_Expectancy_Rate_Country <- sqlQuery(Local, "select Country_Name from Countries,Countries_Metrics
where Life_Expectancy in (select min(Life_Expectancy) from Countries_Metrics) 
                                                and Countries.Country_Code=Countries_Metrics.Country_Code
                                                ")

Life_Expectancy_Rates_Max_Min <- sqlQuery(Local,"select max(Life_Expectancy) as Highest_Life_Expectancy, min(Life_Expectancy) as Lowest_Life_Expectancy, Avg(Life_Expectancy) as Average_Life_Expectancy
from Countries_Metrics")

Life_Expectancy_Highest_lowest <- list()
Life_Expectancy_Highest_lowest[[ "Highest_Country" ]] <- Highest_Life_Expectancy_Rate_Country$Country_Name
Life_Expectancy_Highest_lowest[[ "Lowest_Country" ]] <- Lowest_Life_Expectancy_Rate_Country$Country_Name
Life_Expectancy_Highest_lowest[["Highest_Value"]] <- Life_Expectancy_Rates_Max_Min$Highest_Life_Expectancy
Life_Expectancy_Highest_lowest[["Lowest_Value"]] <- Life_Expectancy_Rates_Max_Min$Lowest_Life_Expectancy
Life_Expectancy_Highest_lowest[["Average Value"]] <- Life_Expectancy_Rates_Max_Min$Average_Life_Expectancy

kable(Life_Expectancy_Rates_Continents)

kable(as.data.frame(Life_Expectancy_Highest_lowest))

Analysis on Government Types in Countries

Type_of_Government <- sqlQuery(Local,"select Government_Type, count(*) as Total_Number_of_Countries 
from Countries group by Government_Type order by Government_Type order by Total_number_of_Countries")

kable(Type_of_Government)

Analysis on the GDP per Capita (World’s Richest and Poorest Countries)

Richest_Country <- sqlQuery(Local,"select Country_Name from Countries,Countries_Metrics
where GDP_per_Capita in (select max(GDP_per_Capita) from Countries_Metrics )
and Countries.Country_Code=Countries_Metrics.Country_Code")

Poorest_Country <- sqlQuery(Local,"select Country_Name from Countries,Countries_Metrics
where GDP_per_Capita in (select min(GDP_per_Capita) from Countries_Metrics )
                            and Countries.Country_Code=Countries_Metrics.Country_Code")

GDP_per_Capita_Max_Min <- sqlQuery(Local,"select max(GDP_per_Capita) as Highest_GDP_per_capita, min(GDP_per_capita) as Lowest_GDP_per_capita, Avg(GDP_per_capita) as Average_GDP_per_capita
from Countries_Metrics")

GDP_per_Capita_Highest_Lowest <- list()
GDP_per_Capita_Highest_Lowest[[ "Richest_Country" ]] <- Richest_Country$Country_Name
GDP_per_Capita_Highest_Lowest[[ "Poorest_Country" ]] <- Poorest_Country$Country_Name
GDP_per_Capita_Highest_Lowest[["Highest_Value"]] <- GDP_per_Capita_Max_Min$Highest_GDP_per_capita
GDP_per_Capita_Highest_Lowest[["Lowest_Value"]] <- GDP_per_Capita_Max_Min$Lowest_GDP_per_capita
GDP_per_Capita_Highest_Lowest[["Average Value"]] <- GDP_per_Capita_Max_Min$Average_GDP_per_capita

kable(as.data.frame(GDP_per_Capita_Highest_Lowest))

Analysis on the Economic Development Status of the countries - Proportion of Developed/Developing and Least Developed Countries within Continents

Continents_by_Economic_Status <- sqlQuery(Local, "select Continents.continent, Economy_Status, round((cast(count(*) as float))/cast(Continent_Totals as float)*100,2) as Proportion  from Continents 
inner join Countries on [Country Name]=Country_Name
                inner join Countries_Metrics on Countries.Country_Code=Countries_Metrics.Country_Code
                inner join (select continent,count(*) as Continent_Totals from Continents group by continent) sub on sub.continent=Continents.continent
                group by Continents.continent, Economy_Status, Continent_Totals
                order by continent")

kable(Continents_by_Economic_Status)

ggplot(Continents_by_Economic_Status, aes(x = continent, y = Proportion,fill=Economy_Status)) +
geom_bar(stat='identity')

Does Health Expenditure and Military Expenditure affect the Life Expectancy Rate?

lm_health_life_expectancy <- lm(Life_Expectancy~Health_Expenditure_per_GDP + Military_Expenditure_per_GDP, data = Countries_Metrics)
summary(lm_health_life_expectancy)

## 
## Call:
## lm(formula = Life_Expectancy ~ Health_Expenditure_per_GDP + Military_Expenditure_per_GDP, 
##     data = Countries_Metrics)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -22.269  -3.148   3.302   6.578  13.880 
## 
## Coefficients:
##                              Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   67.7977     2.4735  27.409   <2e-16 ***
## Health_Expenditure_per_GDP     0.3881     0.2694   1.441    0.152    
## Military_Expenditure_per_GDP   0.2606     0.5757   0.453    0.652    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 9.207 on 134 degrees of freedom
##   (60 observations deleted due to missingness)
## Multiple R-squared:  0.01555,    Adjusted R-squared:  0.0008615 
## F-statistic: 1.059 on 2 and 134 DF,  p-value: 0.3498

The linear regression results tell us that neither Health Expenditure nor Military Expenditure affect the Life Expectancy rate significantly. The below plot tells us the same story.

#### Does Literacy Rate affect Technology Penetration Rate??

lm_literacy_tpr <- lm(Technology_Penetration_rate~Literacy_Rate, data=Countries_Metrics)
summary(lm_literacy_tpr)

## 
## Call:
## lm(formula = Technology_Penetration_rate ~ Literacy_Rate, data = Countries_Metrics)
## 
## Residuals:
##      Min       1Q   Median       3Q      Max 
## -0.42471 -0.15224 -0.03571  0.16559  0.52310 
## 
## Coefficients:
##                 Estimate Std. Error t value Pr(>|t|)    
## (Intercept)   -0.4887271  0.0760268  -6.428 1.06e-09 ***
## Literacy_Rate  0.0093271  0.0008741  10.671  < 2e-16 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 0.2169 on 186 degrees of freedom
##   (9 observations deleted due to missingness)
## Multiple R-squared:  0.3797, Adjusted R-squared:  0.3764 
## F-statistic: 113.9 on 1 and 186 DF,  p-value: < 2.2e-16

The linear regression results tell us that literacy rate does affect the technology penetration rate. However, the coefficient is extremely small. This tells us that there might be other factors affecting the Technology penetration rate that we might be missing.

Below is a plot of how the Technology Penetration Rate and the Literacy Rate affect the GDP per Capita-

#### Do geographical conditions like the coastline/land boundaries affect the GDP_per_Capita?

geography_analysis_data <- sqlQuery(Local, "select Countries_Metrics.Country_Code,Land_Boundaries, Coastline, GDP_per_Capita from Countries_Metrics,Countries_Geography
                                     where Countries_Metrics.Country_Code=Countries_Geography.Country_Code")
lm_geography_gdp <- lm(GDP_per_Capita ~ Land_Boundaries+Coastline, data=geography_analysis_data)
summary(lm_geography_gdp)

## 
## Call:
## lm(formula = GDP_per_Capita ~ Land_Boundaries + Coastline, data = geography_analysis_data)
## 
## Residuals:
##    Min     1Q Median     3Q    Max 
## -23136 -12090  -6052   5912 184808 
## 
## Coefficients:
##                   Estimate Std. Error t value Pr(>|t|)    
## (Intercept)      1.885e+04  1.896e+03   9.942  < 2e-16 ***
## Land_Boundaries -1.278e+00  4.340e-01  -2.946  0.00361 ** 
## Coastline        2.344e-01  9.533e-02   2.459  0.01482 *  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 20870 on 194 degrees of freedom
## Multiple R-squared:  0.0599, Adjusted R-squared:  0.0502 
## F-statistic:  6.18 on 2 and 194 DF,  p-value: 0.002501

The linear regression results find these factors significant, however the coefficients are terribly small to have a significant effect.

Multiple Linear Regression

lm.fit <- lm(GDP_per_Capita ~. -Country_Code -Population -GDP -Labor_Force -Internet_Users -Economy_Status, data=Countries_Metrics)
summary(lm.fit)

## 
## Call:
## lm(formula = GDP_per_Capita ~ . - Country_Code - Population - 
##     GDP - Labor_Force - Internet_Users - Economy_Status, data = Countries_Metrics)
## 
## Residuals:
##      Min       1Q   Median       3Q      Max 
## -14435.9  -4006.1    -50.9   2298.3  27017.7 
## 
## Coefficients:
##                                          Estimate Std. Error t value
## (Intercept)                            -1.256e+04  7.965e+03  -1.577
## Life_Expectancy                        -1.760e+01  1.197e+02  -0.147
## Literacy_Rate                          -3.201e+01  5.407e+01  -0.592
## Health_Expenditure_per_GDP              2.378e+02  2.249e+02   1.057
## Military_Expenditure_per_GDP            1.453e+03  4.917e+02   2.956
## Renewable_Water_Resources              -5.575e-01  6.613e-01  -0.843
## Net_Migration_Rate_per_1000_population  1.199e+02  7.855e+01   1.526
## Commercial_Prime_Lending_Rate          -2.809e+02  1.102e+02  -2.549
## Labor_force_participation_rate          4.180e+04  7.843e+03   5.329
## Technology_Penetration_rate             3.626e+04  4.101e+03   8.842
## Population_Density                      2.750e+00  8.884e-01   3.096
##                                        Pr(>|t|)    
## (Intercept)                             0.11764    
## Life_Expectancy                         0.88336    
## Literacy_Rate                           0.55503    
## Health_Expenditure_per_GDP              0.29265    
## Military_Expenditure_per_GDP            0.00381 ** 
## Renewable_Water_Resources               0.40104    
## Net_Migration_Rate_per_1000_population  0.12974    
## Commercial_Prime_Lending_Rate           0.01218 *  
## Labor_force_participation_rate         5.24e-07 ***
## Technology_Penetration_rate            1.64e-14 ***
## Population_Density                      0.00249 ** 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 6770 on 111 degrees of freedom
##   (75 observations deleted due to missingness)
## Multiple R-squared:  0.8151, Adjusted R-squared:  0.7985 
## F-statistic: 48.94 on 10 and 111 DF,  p-value: < 2.2e-16

The results show us that only the variables Military_Expenditure_per_GDP, Labor_force_participation_rate, Population_Density, Commercial_Prime_Lending_Rate and Technology_Penetration_rate are statistically significant. So we try to fit a linear regression model with only these parameters. By doing this, the R-squared increases from

lm.fit.significant <- lm(GDP_per_Capita ~ Military_Expenditure_per_GDP+Labor_force_participation_rate+Technology_Penetration_rate+Population_Density+Commercial_Prime_Lending_Rate, data=Countries_Metrics)
summary(lm.fit.significant)

## 
## Call:
## lm(formula = GDP_per_Capita ~ Military_Expenditure_per_GDP + 
##     Labor_force_participation_rate + Technology_Penetration_rate + 
##     Population_Density + Commercial_Prime_Lending_Rate, data = Countries_Metrics)
## 
## Residuals:
##      Min       1Q   Median       3Q      Max 
## -14641.1  -3510.5   -587.8   2718.1  28012.3 
## 
## Coefficients:
##                                  Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                    -10596.776   3756.073  -2.821 0.005586 ** 
## Military_Expenditure_per_GDP     1324.081    471.889   2.806 0.005842 ** 
## Labor_force_participation_rate  34662.375   7171.069   4.834 3.95e-06 ***
## Technology_Penetration_rate     34991.865   2851.159  12.273  < 2e-16 ***
## Population_Density                  3.088      0.872   3.541 0.000564 ***
## Commercial_Prime_Lending_Rate    -316.769    106.601  -2.972 0.003570 ** 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 6947 on 122 degrees of freedom
##   (69 observations deleted due to missingness)
## Multiple R-squared:  0.7902, Adjusted R-squared:  0.7816 
## F-statistic: 91.88 on 5 and 122 DF,  p-value: < 2.2e-16

plot(lm.fit.significant)

We see that the R-squared has decreased only slightly even after excluding half of the variables. This is a small cost to pay for greater interpretability of the new model.

Regression Decision Tree

tree_model_gdp <- tree(GDP_per_Capita~. -Country_Code -Population -GDP -Labor_Force -Internet_Users -Economy_Status, data=Countries_Metrics)
summary(tree_model_gdp)

## 
## Regression tree:
## tree(formula = GDP_per_Capita ~ . - Country_Code - Population - 
##     GDP - Labor_Force - Internet_Users - Economy_Status, data = Countries_Metrics)
## Variables actually used in tree construction:
## [1] "Technology_Penetration_rate"           
## [2] "Literacy_Rate"                         
## [3] "Renewable_Water_Resources"             
## [4] "Net_Migration_Rate_per_1000_population"
## [5] "Life_Expectancy"                       
## Number of terminal nodes:  8 
## Residual mean deviance:  29300000 = 3.34e+09 / 114 
## Distribution of residuals:
##     Min.  1st Qu.   Median     Mean  3rd Qu.     Max. 
## -15540.0  -2543.0   -646.5      0.0   2170.0  26670.0

The decision tree model has used the variables Technology_Penetration_rate, Literacy_Rate, Renewable_Water_Resources, Net_Migration_Rate_per_1000_population and Life_Expectancy. This differs from our Linear Regression Model quite a bit in the variables.

What are the factors that affect the Economic Development Status of a country?

We will again answer this question by fitting two different models - the Linear Discriminant Model and the Classification Decision tree Model.

Linear Discriminant Model -

lda.fit <- lda(Economy_Status~. -Country_Code -Population -GDP -Labor_Force -Internet_Users , data=Countries_Metrics)
lda.fit

## Call:
## lda(Economy_Status ~ . - Country_Code - Population - GDP - Labor_Force - 
##     Internet_Users, data = Countries_Metrics)
## 
## Prior probabilities of groups:
##                     Developed                    Developing 
##                     0.4016393                     0.3770492 
## Least Developed Country (LDC) 
##                     0.2213115 
## 
## Group means:
##                               Life_Expectancy Literacy_Rate
## Developed                            77.51143      97.24898
## Developing                           71.48891      88.07609
## Least Developed Country (LDC)        58.94889      60.95185
##                               Health_Expenditure_per_GDP
## Developed                                       7.926531
## Developing                                      6.219565
## Least Developed Country (LDC)                   7.492593
##                               Military_Expenditure_per_GDP
## Developed                                         1.926327
## Developing                                        1.939565
## Least Developed Country (LDC)                     1.768889
##                               Renewable_Water_Resources
## Developed                                      348.2045
## Developing                                     575.4443
## Least Developed Country (LDC)                  206.8944
##                               Net_Migration_Rate_per_1000_population
## Developed                                                  1.9793878
## Developing                                                 0.5406522
## Least Developed Country (LDC)                             -0.8496296
##                               Commercial_Prime_Lending_Rate GDP_per_Capita
## Developed                                          5.806735      31000.451
## Developing                                        12.865000       7586.505
## Least Developed Country (LDC)                     15.888889       1631.816
##                               Labor_force_participation_rate
## Developed                                          0.4869111
## Developing                                         0.4198059
## Least Developed Country (LDC)                      0.3858228
##                               Technology_Penetration_rate
## Developed                                      0.59061088
## Developing                                     0.21598043
## Least Developed Country (LDC)                  0.05317401
##                               Population_Density
## Developed                               327.2058
## Developing                              107.5172
## Least Developed Country (LDC)           142.4166
## 
## Coefficients of linear discriminants:
##                                                  LD1           LD2
## Life_Expectancy                        -2.501811e-02 -1.025452e-01
## Literacy_Rate                          -5.446456e-02 -5.399292e-02
## Health_Expenditure_per_GDP              9.362533e-02  8.551849e-02
## Military_Expenditure_per_GDP            6.116670e-02  5.550107e-02
## Renewable_Water_Resources              -4.687629e-05 -2.550731e-04
## Net_Migration_Rate_per_1000_population  4.352336e-03  8.784090e-03
## Commercial_Prime_Lending_Rate           5.740350e-02  1.809983e-02
## GDP_per_Capita                         -9.208882e-05  5.577265e-05
## Labor_force_participation_rate          1.582394e+00  2.219273e+00
## Technology_Penetration_rate             2.043357e-02  3.075307e+00
## Population_Density                      3.097789e-04  2.648935e-05
## 
## Proportion of trace:
##    LD1    LD2 
## 0.8527 0.1473

Classification Tree Model

tree_model_economy_status <- tree(Economy_Status~. -Country_Code -Population -GDP_per_Capita -Labor_Force -Internet_Users , data=Countries_Metrics)
summary(tree_model_economy_status)

## 
## Classification tree:
## tree(formula = Economy_Status ~ . - Country_Code - Population - 
##     GDP_per_Capita - Labor_Force - Internet_Users, data = Countries_Metrics)
## Variables actually used in tree construction:
## [1] "Technology_Penetration_rate"           
## [2] "Life_Expectancy"                       
## [3] "Literacy_Rate"                         
## [4] "Population_Density"                    
## [5] "Net_Migration_Rate_per_1000_population"
## [6] "Commercial_Prime_Lending_Rate"         
## [7] "GDP"                                   
## Number of terminal nodes:  12 
## Residual mean deviance:  0.5317 = 58.48 / 110 
## Misclassification error rate: 0.1066 = 13 / 122

The Classification Tree model has deemed the below variables important to determine the Economy Status of a country. [1] “Technology_Penetration_rate” “Life_Expectancy”
[3] “Literacy_Rate” “Population_Density”
[5] “Net_Migration_Rate_per_1000_population” “Commercial_Prime_Lending_Rate”
[7] “GDP”