GitHub - DiarCode/michi: DTS-GSSF is a dual-timescale graph state-space forecasting framework for real-time passenger flow that combines long-memory spatio-temporal modeling with online residual correction, drift-triggered low-rank adaptation, and hierarchical reconciliation for coherent station→line→network predictions.

Branches Tags

Name		Name	Last commit message	Last commit date
Latest commit History 3 Commits
checkpoints		checkpoints
data		data
implementations		implementations
model_eval_q1_full		model_eval_q1_full
outputs		outputs
researches		researches
.DS_Store		.DS_Store
.gitignore		.gitignore
.python-version		.python-version
MODEL_ARCHITECTURE.md		MODEL_ARCHITECTURE.md
README.md		README.md
main.py		main.py
main.tex		main.tex
model_evaluation.py		model_evaluation.py
pyproject.toml		pyproject.toml
uv.lock		uv.lock

Repository files navigation

About

DTS-GSSF is a dual-timescale graph state-space forecasting framework for real-time passenger flow that combines long-memory spatio-temporal modeling with online residual correction, drift-triggered low-rank adaptation, and hierarchical reconciliation for coherent station→line→network predictions.