CS385

This is repo recording my solution to two course project in SJTU CS385 (上海交通大学机器学习课程)

Project1

In project1, I inplemented several machine learning methods through SVHN including

Basic Linear Model
- Logistic Regression
- Lasso Regression
- Ridge Regression
- Spline Regression
Linear Analysis Mode
- LDA
- GNM
SVM
- linear kernel
- polynomial kernel
- sigmoid kernel
- RBF kernel
Neural Networks
- MobileNetV3
- NiN
- LSTM
- MLP You can refer to project1 folder which including the source code (partly handwrote) to learn some concepts about machine learning. I also give some experience and conclusion from completing this porject in my report.

Project2

In project2 , our group gives a deep insight on various machine learing models and tasks. I am resposible for multi-category classfication tasks over CIFAR10 and Fashion-MNIST. The implemented models include:

AlexNet
Vgg family
ResNet family

You can refer to project2 folder which including the source code (partly handwrote) to learn some concepts about machine learning. The link to my teammates including CelebA problem, VAE problems, GAN problems and LSUN problems.

Visualization

Part of the codes I used for visualization (GradCAM, guided backpropagation, TSNE, PCA) is provided. These are just example codes.

Ackowledgements

I give my sincerest appreciation to Prof. Zhang, Dr. Cheng and Dr. Ren for giving me this chance to contribute my efforts in the work of machine learning. I also thank them for their generous help.

Contact