We started looking at the new dataset today. It comes from ACLED and it has information about political violence and other sorts of violence. Gary also talked about “Clustering” specifically “K-means clustering” which is a very popular technique used in unsupervised learning. which are used to group similar type of data which have no labels.
After looking at the ACLED dataset I have come up with two questions..
- Do protests turn violent in certain regions? or do certain types of protests by certain people always become violent
- Do certain actors appear frequently in violent events?