Machine Learning 강의노트 01. Introduction 02. Linear Regression 1) Univariate Linear Regression a. Pytorch Example b. TensorFlow Example 2) Parameter Learning (Gradient Descent) a. Pytorch example b. TensorFlow example 3) Multivariate Linear Regression a. Pytorch Example 4) Gradient Descent in Practice a. Pytorch Example: Dataset 5) Features and Polynomial Regression a. Pytorch Example: Evaluation 6) Computing Parameters Analytically 03. Logistic Regression 1) Hypothesis Representation 2) Decision Boundary 3) Cost Function & Gradient Descent 4) Advanced Optimization 5) Multi-Class Classification 04. Regularization 1) Cost Function 2) Regularized Linear Regression 3) Regularized Logistic Regression 05. Neural Networks 1) Model Representation 2) Cost Function 3) Back-Propagation Algorithm I 4) Back-Propagation Algorithm II 5) Implementation Notes 6) Put it Together 06. Advice for Applying Machine Learning 1) Machine Learning Diagnostic 2) Bias vs. Variance 3) Learning Curves 4) Deciding What to do Next (Revisited) 07. Machine Learning System Design 1) Building a Spam Classifier 2) Handling Skewed Data 3) Using Large Data Sets 08. Support Vector Machines 1) Decision Boundary: Large Margin 2) Kernels 3) Using an SVM 09. Unsupervised Learning 1) K-Means Algorithm 2) Optimization Objective 3) Random Initialization 4) Choosing the Number of Clusters 10. Dimensionality Reduction 1) Principal Component Analysis 2) Reconstruction from Compressed Representation 3) Choosing the Number of Principal Components 4) Advice for Applying PCA 5) Factor Analysis 6) Linear Discriminant Analysis 11. Anomaly Detection 1) Density Estimation 2) Building an Anomaly Detection System 3) Choosing what Features to Use 12. Recommender Systems 1) Content Based Recommendation 2) Collaborative Filtering 3) Low Rank Matrix Factorization 4) Implementational detail 13. Large Scale Machine Learning 1) Gradient Descent with Large Datasets 2) Advanced Topics 14. Application Example: Photo OCR A-1. Probability and Likelihood A-2. AdaBoost A-3. Gaussian Mixture Model A-4. Feature Selection
Published with WikiDocs