Financial News Sentiment Analysis
Sentiment classifier for financial headlines using Naive Bayes, FFNs, and BERT-based RNNs. F1 0.773 multiclass / 0.892 binary.
AIMLNLPDeep LearningPythonFinanceFinTechSentiment Analysis
Jupyter Notebook Started Jun 16, 2025 Updated Jun 26, 2025
Built a sentiment classifier for financial news headlines using Naive Bayes, Feedforward Neural Networks and BERT-based RNNs on a domain-labeled dataset (~5K headlines). Tackled class imbalance and improved performance through lemmatization and optimized vectorization.
Achieved F1-score of 0.773 (multiclass) and 0.892 (binary) with neural networks. Also explored semantic similarity with PMI and GloVe.