Research and development - Seminars
Collecting data on human rights violations in the context of conflict is difficult and dangerous, and the resulting data is often incomplete. Some victims' stories are never documented and those whose stories are documented may be missing key information. In addition, the data that are documented are not necessarily statistically representative of the entire victim population. Drawing population-level inferences from these data without correcting for missing data risks incorrectly answering questions about patterns of violence. This talk will introduce the methodology behind statistics on homicide, enforced disappearance, kidnapping, illicit recruitment and displacement in the context of the Colombian conflict. It will discuss the methods that were used to deduplicate over 100 databases, fill in missing fields in the documented data, and estimate the total size of the victim population, including those victims that were never documented by any source.
YouTube – Quantil Matemáticas Aplicadas
Not available
Get information about Data Science, Artificial Intelligence, Machine Learning and more.