Skip to content

Physics-Informed LSTM: Deep learning architecture incorporating stochastic car-following structure and LSTM for predicting human driving behavior (Universal Differential Equations)

License

Notifications You must be signed in to change notification settings

Glodanale/PILSTM

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

6 Commits
 
 
 
 
 
 
 
 

Repository files navigation

PILSTM

Physics-Informed LSTM: This workflow follows the Universal Differential Equation methodology via the incorporation of stochastic modeling, the Intelligent Driving Model, with an LSTM structure in a joint-training paradigm for predicting human driving behavior.

Note: This repository reflects the workflow of Alex Glover's Master of Science in Computer Science: Machine Learning and AI thesis at East Tennessee State University, accepted March of 2025

Variations

Sequence Handling: Truncated sequences, Padded & Masked sequences

Intelligent Driving Model: Linear, Nonlinear

Experimental Factors

LSTM Cell Count

Float Precision

Random Seed Value: NumPy, Scikit-Learn, TensorFlow

Experimental Cycles

Training-Only Behaviors

Training-Validation Behaviors

Testing Final Values

About

Physics-Informed LSTM: Deep learning architecture incorporating stochastic car-following structure and LSTM for predicting human driving behavior (Universal Differential Equations)

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published