ckswjd99-at-snu / machinelearning-2022-1

SNU ECE Machine Learning Fundamentals & Applications (M2608.001300) - 2022 Spring

Machine Learning 2022-1

Assignments and Final Project in SNU ECE Machine Learning Fundamentals & Applications (M2608.001300) lecture, at 2022 spring.

HW1: Logistic Regression

• Training logistic regression with scikit-learn library
• To implement logistic regression with numpy (from scratch)
• Logistic regression with regularization
• Multi-class classification with logistic regression

HW2: Support Vector Machine

Report

• Hard margin SVM
• Soft margin SVM
• Non-linear classification with kernel trick

HW3: Convolutional Neural Network

Report / Dataset

• Image classification model design and training
• Combining GoogLeNet and ResNet method
• 72% accuracy on CIFAR-10 dataset

HW4

Dataset

Pt1: Hidden Markov Model

• Word generation model
• Implement HMM forward algorithm
• Implement Viterbi algorithm

Pt2: Recurrent Neural Network

• Word generation model
• Implement RNN model

Final Project

Report / Easy Dataset / Normal Dataset

• Recognizing letter images
• Predicting next letter of a letter sequence
• Combining VGG and LSTM method
• 98.89% accuracy on the easy dataset
• 96.15% accuracy on the normal dataset