HKU Engineering FYP24099

Sentiment Analysis on Cryptocurrency using Machine Learning and NLP

Sentiment Analysis on Cryptocurrency using Machine Learning and NLP

SUPERVISOR: LIU QI

Group: Ezen CHONG, Gillian Ru Qu TOO, Keith Kai Xuan CHAN, Zi Zheng FONG HU

Building exterior in Toronto, Canada

Can new AI models be used to predict Cryptocurrency Market movements?

Current Progress

Data Preprocessing

  • Initiate data collection
  • Utilize open-sourced models like LLAMA
  • import models locally via techniques such as QLORA
Tourist taking photo of a building
Windows of a building in Nuremberg, Germany

Machine Learning Model Training

  • Train ML models (Logistic Regression, SVM) for sentiment classification
  •  Develop sentiment indices for cryptocurrencies
  • Develop predictive models for price movements and evaluate model performance

Refinement and Optimization

  • Optimize models based on evaluation results
  • Perform extensive back testing with a larger dataset
  • Analyze model robustness and reliability
Tourist taking photo of a building