Yusen Liu's Projects
This project addresses binary and multiclass classification problems using both XGBoost and neural networks, starting from basic structures and expanding to deeper architectures like CNNs, VGG, and EfficientNet V2.
Implementation of SRResnet and SRGAN models on the DIV2K dataset. This project focuses on enhancing image resolution using state-of-the-art techniques, with certain hyperparameter modifications and gradient clipping to address computing power limitations and stabilize training.
Code repository for the O'Reilly book 'Cloud Native Data Center Networking'
This GitHub project focuses on sentiment analysis for Twitter data using RoBERTa and traditional algorithms. We preprocess the datasets with text cleaning and tokenization. RoBERTa's contextualized word embeddings and fine-tuned architecture excel in sentiment classification, outperforming traditional algorithms, especially with larger datasets.
12 weeks, 26 lessons, 52 quizzes, classic Machine Learning for all