paper-conference

Leveraged do-calculus-based causal learning algorithms on top of existing transformer-based vulnerability detection models to systematically eliminate reliance on spurious features, enhancing causal-based predictions.

Surveyed and reproduced 9 state-of-the-art DL models on the Devign and MSR datasets. Investigated model capabilities, training data effects, and interpretability, revealing key insights to improve model robustness and understandability.

The paper addresses the challenge of traffic congestion research in developing countries, proposing a cost-effective method of collecting data via Google Map’s traffic layer. It demonstrates the effectiveness of this data in predicting traffic congestion using established models and highlights distinct traffic patterns between weekdays and weekends.