Skip to content

Latest commit

 

History

History
11 lines (5 loc) · 497 Bytes

README.md

File metadata and controls

11 lines (5 loc) · 497 Bytes

twitter-sentiment-analysis

Explored the effect of different BERT-based models on twitter sentiment 3-class classification.

Proposed to detect the relevancy between keywords and corresponding sentence sentiment by combining self-attentions and BERT NER.

Explored the relationship between twitter sentiment of Tesla and its stock price/return.

Explore of different hyper-parameter tuning methods for transformers: random search, bayesian optimization and population based training.