Nutrition Analysis of Stunting in Depok, Indonesia using R

Methods: Multivariate Adaptive Regression Splines (MARS)

\

Analysis Purpose

Conference and Publication of Journal

Data Sources

Secondary Data from Public Health, University of Indonesia (UI)

Sample Size

132 samples (Minimal Data for MARS Methods: 100)

List of Variables

y = Status
x1 = Power
x2 = Carbohydrate
x3 = Protein
x4 = Fat
x5 = Vitamin A
x6 = Carotene
x7 = Iron
x8 = Calcium
x9 = Zinc
x10 = Vitamin D
x11 = Vitamin C

Descriptive Statistics

	Min	Max	Median	Mean
Power	256	6750	1037	1185.4
Carbohydrate	13	820.5	135.15	153.79
Protein	5.1	192.6	33.4	38.14
Fat	2	841.1	44.7	57.59
Vitamin A	94.9	9794	855	1231.7
Carotene	0	329	0	5.37
Iron	0	99.8	9.05	12.61
Calcium	14	5744.6	364.9	589
Zinc	0.6	31.7	4.65	5.559
Vitamin D	0	1000	4	29.35
Vitamin C	3	820	74.5	103.8

Step of Analysis

1. Preprocessing Data (Manual)
2. Reliable Testing
3. MARS Modeling
4. Linear Regression Modeling & Testing
5. MARS Importance Variable

Final Output

Variable Importance of Nutrition

	N subsets	GCV	SSE
Power	8	100	100
Zinc	7	47.6	49.3
Vitamin A	6	41	42.1
Fat	1	7.9	9.4

Dataset [! SOON]

1. Original Dataset
2. Linear Regression Dataset

Teams

Afifa Ayu Widhiyanthi
Bryan Alfason Sunjaya
I Gede Arya Krisnadi
Naufal Tansel

Code by afifadayu