THE ROLE OF SINGULAR VALUE DECOMPOSITION IN CORRESPONDENCE ANALYSIS MAPPING: CASE STUDY OF STUNTING DATA FROM LEBAK AND PANDEGLANG IN 2023
Correspondence Analysis a multivariate analysis method used to map relationships between qualitative variables in a low-dimensional space. The fundamental concept of correspondence analysis is the row profiles and column profiles where the columns represent the position vectors of the observed c...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Project |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/84067 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | Correspondence Analysis a multivariate analysis method used to map
relationships between qualitative variables in a low-dimensional space. The
fundamental concept of correspondence analysis is the row profiles and column
profiles where the columns represent the position vectors of the observed
categories. However, this visualization is only applicable to contingency tables
with no more than three categories. Therefore, this final project uses Singular
Value Decomposition for dimensionality reduction to obtain a visualization in the
form of correspondence analysis maps. This approach is applied to stunting data
collected by the Indonesian Population Coalition in 2023, focusing on variables
such as the district of residence, fever management methods, educational level of
caregivers, and sources of information on stunting. The analysis reveals
significant associations among these variables, providing insights that could
inform public health strategies.
|
|---|