Research Article
3D Model Classification Based on Bayesian Classifier with AdaBoost
Algorithm 1
The construction of Classifier(X) based on Adaboost.
| Input: training instances (X1, y1), (X2, y2), …, (Xn, yn). | | Output: strong classifier Classifier(X). | | Step 1. Initialize weight 1i of training instances, and iteration time K: | | | | Step 2. for (k = 1; k ≤ K; k++){ | | Step 2.1 Use training instances with weight k1, k2, …, kn to train classifier Classifierk(X). | | Step 2.2 Calculate error rate ek of Classifierk(X): | | | | where I (Classifierk(Xi)≠yi) is 0 or 1. If Classifierk(Xi)≠yi, then I(Classifierk(Xi)≠yi) = 1. Otherwise, I(Classifierk(Xi)≠yi) = 0. | | Step 2.3 Calculate weight coefficient αk of Classifierk(X): | | | | Step 2.4 Update weight k + 1,i: | | | | | | } | | Step 3. Combine Classifierk(X) according to weight coefficient αk, (k = 1, 2, …, K): | | |
|