web数据挖掘__4监督学习2.ppt课案.pptx

Supervised Learning (2);Road Map;Bayesian Theorem: Basics;Bayesian Theorem: Basics;Bayesian Theorem: Basics;Na?ve Bayes Classifier;Na?ve Bayes Classifier;Conditional independence assumption ;Final na?ve Bayesian classifier;Classify a test instance;An example;An Example (cont …);Training dataset;Na?ve Bayesian Classifier: An Example;Additional issues ;Avoiding the 0-Probability Problem;On na?ve Bayesian classifier;Road Map;Introduction;Basic concepts;The hyperplane;Maximal margin hyperplane;23;24;Linear SVM: separable case;Compute the margin;Compute the margin (cont …);A optimization problem!;Solve the constrained minimization;Kuhn-Tucker conditions;Solve the problem;The final decision boundary;Linear SVM: Non-separable case;Relax the constraints ;Geometric interpretation;Penalize errors in objective function;New optimization problem;Kuhn-Tucker conditions ;The final decision boundary;How to deal with nonlinear separation?;41;Space transformation;Geometric interpretation;Optimization problem becomes ;An example space transformation;Problem with explicit transformation;Kernel functions;An example kernel function;Kernel trick;Commonly used kernels;Some other issues in SVM;多类的情况 ;SVM方法的特点;SVM特点;SVM方法的特点;Road Map;近邻分类方法;k-Nearest Neighbor Classification (kNN);kNNAlgorithm;Example: k=6 (6NN);最近邻法;62;k-近邻法;Parameter selection;K-effect;Parameter selection;kNN复杂度;Discussions;Road Map;Combining classifiers;Bagging;Bagging (cont…);Bagging (cont …);Summary