Lecture slides

• Lecture 1 – Overview: 运筹学和最优化方法 [posted on March 5]

• Lecture 2 – 向量，矩阵，子空间 [posted on March 12]

• Lecture 3 – 线性规划的基本概念与理论 [posted on March 19】

• Lecture 4 – 线性规划的标准形式 [posted on Mar 19】

• Lecture 5 – 凸函数 [posted on Apr 2】

• Lecture 6 – 无约束条件的凸优化一 [posted on Apr 23]

• Lecture 7 – 无约束条件的凸优化二 [posted on May 9]

• Lecture 8 – 无约束条件的凸优化三 [posted on May 14]

• Lecture 9 – 支持向量机(SVM) [posted on Apr 16]

• Lecture 10 – 含约束条件的凸优化一 [posted on May 16]

• Lecture 11 – 含约束条件下的对偶理论 [posted on May 21]

• Lecture 12 – Introduction to Neural Networks [posted on June 11]

• Lecture 13 – Online Optimization [posted on June 13]

Homework

• Homework 1

• Homework 2

• Homework 3

Project

• Handwritten Digits Classification. Dataset is available on THE MNIST DATABASE. In this project, you have two options:

SVM Methods

1. Apply the basic SVM model to train the training set and classify the digits in the test set.
2. Use kernel method to retrain the SVM model and classify the digits.

• 报告撰写范例

CNN Methods:

Apply the Convolution Neural Network method to perform the training and testing.