A hybrid Bi-LSTM plus LightGBM model with physics-inspired handcrafted features is claimed to improve lane-change intention prediction on highD and exiD, but the reported gains are compromised by test-set leakage and inconsistent results.
Factors affecting lane change crashes,
1 Pith paper cite this work, alongside 91 external citations. Polarity classification is still indexing.
1
Pith paper citing it
91
external citations · OpenAlex
fields
cs.LG 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Evolutionary Physics-Informed Temporal Fusion for Lane-Change Intention Prediction
A hybrid Bi-LSTM plus LightGBM model with physics-inspired handcrafted features is claimed to improve lane-change intention prediction on highD and exiD, but the reported gains are compromised by test-set leakage and inconsistent results.