Data Processing for an International Humanitarian NGO

Data Processing for an international humanitarian NGO to classify the countries using socio-economic and health factors that determines the overall development of a country and to also suggest the countries that have to given utmost importance.

Methodology:

This analysis and prediction can be done by following the steps below:

1. Analysing the dataset
2. Performing Exploratory Data Analysis
3. Plotting Correlation Graphs to analyse the data
4. Scaling the data
5. Performing Principal Component Analysis
6. K-Means Clustering of data
7. Hierarchical Clustering of data
8. Comparing the results and predicting the countries having dire need of aid.

Conclusion:

Choosing K-Means Clustering Algorithm over Hierarchical Clustering Algorithm because the clusterID value counts were properly divided and visualizing each cluster was possible.

The countries in dire need of aid by K-Means (17) were more than by Hierarchical Clustering(5)

After comparing both K-means and Hierarchical clustering method, we can go with the K-means outcomes as the plots are clearly visible. As in both the methods, the produced mean for underdeveloped countries were almost same. i.e. the deciding number of clusters as 3 was profitable.

1. After grouping all the countries into 3 groups by using some socio-economic and health factors, we can determine the overall development of the countries.

2. Here, the countries are categorised into list of developed countries, developing countries and under-developed countries.

3. In Developed countries, we can see the GDP per capita and income is high where as Death of children under 5 years of age per 1000 live births i.e. child-mort is very low, which is expected.

4. In Developing countries and Under-developed countries, the GDP per capita and income are low and child-mort is high. Specially, for under-developed countries, the death rate of children is very high.

Recommendations:

1. From bar chats, we can clearly see the socio-economic and heath situation of the under developed countries.

2. In countries like Haiti, Sierra Leone, Chad, etc., the death rate of children under 5 years of age per 1000 (child-mort) is high.

3. Countries like Burundi, Congo, Niger, etc., GDP per capita is very low & the income per person is also low. So, these countries are considered as Poor countries.

4. If Child Mortality Rate is decreased and GDPP , Income is increased in Under-developed countries, the need will be resolved.