本仓库内为2022年春季学期中山大学最优化理论课程的期末大作业。
文件说明:
- 实验报告.pdf 为实验报告,包含算法设计、实验结果等内容;
- task1_1.py 对应问题1的邻近点梯度法;
- task1_2.py 对应问题1的交替方向乘子法;
- task1_3.py 对应问题1的次梯度法;
- task2_1GD.py 对应问题2的梯度下降法;
- task2_2SG.py 对应问题2的随机梯度法;
- Optimization-2022 为本次期末大作业要求:
- task 1 为使用不同算法求解一范数规范化最小二乘模型;
- task 2 为使用不同算法求解 MNIST 数据集上的分类问题并比较对应的精度 ,同时回答相应问题
注:mnist数据集通过源代码可以直接下载,因此不附在文件夹中。
This is the final assignment for the Optimization Theory course of Sun Yat-sen University in the spring semester of 2022.
Document description:
-
Experiment report.pdf is the experiment report, including algorithm design, experiment results and other content;
-
task1_1.py corresponds to the gradient method of neighboring points in problem 1;
-
task1_2.py corresponds to the alternate direction multiplier method of problem 1;
-
task1_3.py corresponds to the sub-gradient method of problem 1;
-
task2_1GD.py corresponds to the gradient descent method of problem 2;
-
task2_2SG.py corresponds to random gradient method in problem 2;
-
Optimization-2022 is the final assignment requirement:
- task 1 is to solve the one-norm normalized least squares model using different algorithms;
- task 2 is to use different algorithms to solve classification problems on MNIST dataset, compare the corresponding accuracy, and answer the corresponding questions.
Note: The MNIST dataset can be downloaded directly from the source code, so it is not attached to the folder.