Table 7 Information gain values for different clusters using cdp-method [5] and the proposed d-method
From: Enhancing correlated big data privacy using differential privacy and machine learning
Cluster no. | Dataset a | Dataset b | Dataset c | |||
|---|---|---|---|---|---|---|
Info gain values using cdp-method | Info gain values using d-method | Info gain values using cdp-method | Info gain values using d-method | Info gain values using cdp-method | Info gain values using d-method | |
Cluster 1 | 0.0 | 0.163 | 0.038 | 0.144 | 0.019 | 0.078 |
Cluster 2 | 0.020 | 1.0 | 0.055 | 0.312 | 0.176 | 0.221 |
Cluster 3 | 0.026 | 0.149 | 0.070 | 0.971 | 0.036 | 0.349 |
Cluster 4 | 0.039 | 0.731 | 0.085 | 0.712 | 0.069 | 0.918 |
Cluster 5 | 0.084 | 0.096 | 0.091 | 0.196 | 0.088 | 0.159 |