Topological Data Analysis and UNICEF Multiple Indicator Cluster Surveys

Multiple Indicator Cluster Surveys (MICS), supported by UNICEF, are one of the most important global household survey programs that provide data on health and education women children. We analyze Serbia 2014–2015 MICS dataset using topological analysis which treats cloud as a space extracts information about its intrinsic geometric properties. In particular, our uses Mapper algorithm, dimension-reduction clustering method produces graph from cloud. The resulting provides insight into various relationships between wealth—as expressed wealth index, an indicator extracted data—and other parameters such urban/rural setting, ownership items, prioritization possessions. Among uses, these findings can serve to inform policy providing hierarchy essential amenities. They also potentially be used refine index or deepen understanding what it captures.