Table 1 Decay values, α, for each of the original data dimensions and the dimension reduced models governing the decay of consumption deprivation, CD(y), in Fig. 2.

From: A social engineering model for poverty alleviation

Data/Model

CD(y) Decay rate, α + 2

Cereals

3.5663

Other food

1.4853

Non-food

0.81851

Multivariate-NeuroScale

1.0739

Multivariate-LLE

0.79709

Multivariate-IsoMap

1.0425

Multivariate-CCA

0.95466

Multivariate-PCA

0.95739

  1. The global mappings of IsoMap and NeuroScale show an expansion in the decay in contrast to the multivariate exponents calculated using LLE, CCA, and PCA. Our results show that the multivariate exponent value (α + 2), averaged over all five dimensional reduction measures, is ~1.0, as opposed to ~3.5 for a cereal-consumption-based exponent21.
Back to article page