Measuring Sustainable Development Efficiency of Urban Logistics Industry

Algorithm 1

Efficiency evaluation algorithm.

Input: Feature vectors , classification vectors and the size for training set;

Feature vectors , and the size for test set.

The number of classifications and features are and , respectively.

Output: Predicted classifications , efficiency values and ranking list.

1: for

2:   is added into the classification set according to

3: end for

4: for

5:   Compute the classification probability = size;

6:    for

7:   Compute the mean value according to Eq. (2);

8:   Compute the variance according to Eq. (3);

9:     end for

10: end for

11: for

12:  for

13:   for

14:   Compute conditional probability according to Eq. (4);

15:   end for

16:    Compute joint probability using Eq. (5)

17:  end for

18:  Obtain the predicted level c using Eq. (6)

19:  Compute using Eq. (7)

20: end for

21: Return a ranking list according to .