Table 1 Sensitivity of country classification to Temporal weighting (λ).

From: Understanding food loss patterns across developed and developing countries using a GDP, growth rate, and health expenditure-based typology

Lambda

Silhouette Score

Suggested optimal k

Silhouette for suggested optimal k

Developed (n)

Developing (n)

0.00

0.6849006

2

0.6849006

12

93

0.05

0.6819273

2

0.6819273

13

92

0.10

0.6865926

2

0.6865926

13

92

0.15

0.6924942

2

0.6924942

13

92

0.20

0.6987467

2

0.6987467

13

92

0.25

0.6801862

3

0.7161577

13

92

0.30

0.6818329

3

0.7212277

13

92

0.35

0.6877317

3

0.7246317

13

92

0.40

0.687781

3

0.7224686

14

91

0.45

0.6880869

3

0.7251968

14

91

0.50

0.694156

3

0.7270374

14

91

Lambda

Agreement Rate

Countries Changed*

Developed Changed to Developing*

Developing Changed to Developed*

0.00

0.99

1

1

0

0.05

1.00

0

0

0

0.10

1.00

0

0

0

0.15

1.00

0

0

0

0.25

1.00

0

0

0

0.30

1.00

0

0

0

0.35

1.00

0

0

0

0.40

0.99

1

0

1

0.45

0.99

1

0

1

0.50

0.99

1

0

1

Country

λ = 0

λ = 0.05–0.35

λ = 0.4–0.5

Economic Rationale

New Zealand

Developing

Developed

Developed

High GDP; moderate health spending affects weighting

Republic of Korea

Developing

Developing

Developed

Recent rapid growth accentuated at higher λ

All Others

Stable

Stable

Stable

Robust classification across weighting range

  1. * Reclassification relative to λ = 0.20 (highest silhouette score at k = 2).