Explorative and time series analysis of world debt dataset
The dataset conatins unbalanced data of GDP, public debt and GDP/DEBT ration for 187 countries. The time series for each country spans from 1800 to 2020 but each country data dependes on their date of independence and data availability.
This project consists in:
- Domain analysis
- Statistical analysis
- Missing data removal and inference
- Outlier removal
- Clustering
- Multidimensional KMeans
- Agglomerative
- DBSCAN
- KMeans and Agglomerative on Haar wavelet transform
- KMeans and Agglomerative on Dinamic Time Warping data
- Pattern mining
- Apriori on cluster data
More details in the docs