Global Optimization Ensemble Model for Classification Methods

Table 9

Results of adult income dataset: comparison of optimized classification accuracy using GMC model with simple classification using different classifiers.

Algorithm	Classification accuracy	Optimized classification accuracy	Improvement %
K-NN	76.70%	83.20%	6.5%
Decision tree	80.00%	82.20%	2.20%
ID3	75.60%	78.60%	3%
W-PART	81.00%	83.50%	2.4%
W-Prism	81.10%	82.20%	1.1%
W-AODE	80.80%	82.60%	1.8%
W-J48	81.50%	83.00%	1.5%
Random forest	76.10%	77.30%	1.2%
Logistic regression	79.00%	80.00%	1%