Forecasting Without Full Identification: Reliability and Computational Cost in Clot-Growth Simulation

Abstract:We forecast clot growth from sparse early measurements with incomplete biochemical information. Using 1,000 simulated trajectories, we compare inverse search, which weights candidate inputs by their agreement with early observations, with a direct multilayer perceptron (MLP) predicting future clot size. Inverse search is more accurate when tissue factor alone, or tissue factor and fibrinogen, are unknown. When all five inputs are unknown, direct prediction achieves 2.16% pooled relative $L_2$ forecast error, outperforming search with 8,192 candidates using a forward MLP (3.25%) or latent neural ordinary differential equation model (3.36%). A conditional error bound separates candidate disagreement, weight on poor early fits, surrogate error, and candidate coverage. Increasing randomized quasi-Monte Carlo candidate budgets from 512 to 32,768 reduces numerical variation. A covariance score ranks additional assays by expected forecast value while accounting for measurement noise. At 10% range-scaled assay noise, incorporating assay uncertainty reduces forecast error from about 5% to below 3% for both surrogates. An independent reaction cascade verifies candidate integration against tensor quadrature, with mean normalized forecast-integration error below $10^{-5}$ at 32,768 candidates. These results demonstrate accurate clot-growth forecasting without full biochemical identification and support selecting prediction strategies and additional measurements according to accuracy, reliability, and computational cost.

Machine Learning (cs.LG); Quantitative Methods (q-bio.QM); Tissues and Organs (q-bio.TO)

I.2.6; I.6.5; J.3

arXiv:2608.08165 [cs.LG]

(or arXiv:2608.08165v2 [cs.LG] for this version)

https://doi.org/10.48550/arXiv.2608.08165 Focus to learn more arXiv-issued DOI via DataCite

MATCOM_code_only.sha256.txt

clotonet_code/MANIFEST.json

clotonet_code/README.md

clotonet_code/inverse-final.py

clotonet_code/inverse-v3sndde.py

clotonet_code/inverse-v7.py

clotonet_code/inverse-v8.py

clotonet_code/inverse-v9.py

clotonet_code/requirements_reproduction.txt

clotonet_code/revision_amm/NOISY_FACTOR_PROTOCOL.md

clotonet_code/revision_amm/__init__.py

clotonet_code/revision_amm/audit_reproducibility.py

clotonet_code/revision_amm/audit_statistics.py

clotonet_code/revision_amm/check_snode_time_steps.py

clotonet_code/revision_amm/core.py

clotonet_code/revision_amm/export_corpus_svg.py

clotonet_code/revision_amm/export_noisy_factors.py

clotonet_code/revision_amm/export_results.py

clotonet_code/revision_amm/figure_ambiguity.py

clotonet_code/revision_amm/figure_assay.py

clotonet_code/revision_amm/figure_robustness.py

clotonet_code/revision_amm/make_figures.py

clotonet_code/revision_amm/noisy_assays.py

clotonet_code/revision_amm/prepare_references.py

clotonet_code/revision_amm/run_acquisition.py

clotonet_code/revision_amm/run_candidate_stability.py

clotonet_code/revision_amm/run_experiments.py

clotonet_code/revision_amm/run_extrapolation.py

clotonet_code/revision_amm/run_noisy_factors.py

clotonet_code/revision_amm/run_snode_stability.py

clotonet_code/revision_amm/svg_to_manuscript_pdf.py

clotonet_code/revision_amm/test_acquisition.py

clotonet_code/revision_amm/test_candidate_stability.py

clotonet_code/revision_amm/test_candidate_validation.py

clotonet_code/revision_amm/test_core.py

clotonet_code/revision_amm/test_figure_ambiguity.py

clotonet_code/revision_amm/test_figure_assay.py

clotonet_code/revision_amm/test_figure_robustness.py

clotonet_code/revision_amm/test_forecast_framework.py

clotonet_code/revision_amm/test_lily_math.py

clotonet_code/revision_amm/test_noisy_assays.py

clotonet_code/revision_amm/test_noisy_factor_metrics.py

clotonet_code/revision_amm/test_noisy_factor_runner.py

clotonet_code/revision_amm/test_reproducibility.py

clotonet_code/revision_amm/test_snode_stability.py

clotonet_code/revision_amm/test_snode_stability_validation.py

clotonet_code/revision_amm/test_snode_time_steps.py

clotonet_code/revision_amm/test_statistics.py

clotonet_code/revision_amm/theory/ITERATION_2_PROTOCOL.md

clotonet_code/revision_amm/theory/ITERATION_3_PROTOCOL.md

clotonet_code/revision_amm/theory/PROTOCOL.md

clotonet_code/revision_amm/theory/attenuation.py

clotonet_code/revision_amm/theory/export_results.py

clotonet_code/revision_amm/theory/front.py

clotonet_code/revision_amm/theory/make_figures.py

clotonet_code/revision_amm/theory/relaxation.py

clotonet_code/revision_amm/theory/run_attenuation.py

clotonet_code/revision_amm/theory/run_relaxation.py

clotonet_code/revision_amm/theory/run_validation.py

clotonet_code/revision_amm/theory/test_additional.py

clotonet_code/revision_amm/theory/test_attenuation.py

clotonet_code/revision_amm/theory/test_attenuation_runner.py

clotonet_code/revision_amm/theory/test_front.py

clotonet_code/revision_amm/theory/test_relaxation.py

clotonet_code/revision_amm/theory/test_validation.py

clotonet_code/revision_amm/theory/validate_results.py

clotonet_code/revision_amm/validate_candidate_stability.py

clotonet_code/revision_amm/validate_snode_stability.py

clotonet_code/revision_matcom/__init__.py

clotonet_code/revision_matcom/adaptive_assays.py

clotonet_code/revision_matcom/diagnostics.py

clotonet_code/revision_matcom/diagnostics_snode.py

clotonet_code/revision_matcom/direct_baseline.py

clotonet_code/revision_matcom/export_adaptive_assays.py

clotonet_code/revision_matcom/export_diagnostics.py

clotonet_code/revision_matcom/export_direct_matched_cost.py

clotonet_code/revision_matcom/export_factor_noise.py

clotonet_code/revision_matcom/export_forecast_examples.py

clotonet_code/revision_matcom/manufactured.py

clotonet_code/revision_matcom/manufactured_presentation.py

clotonet_code/revision_matcom/package_submission.py

clotonet_code/revision_matcom/test_adaptive_assays.py

clotonet_code/revision_matcom/test_diagnostics.py

clotonet_code/revision_matcom/test_direct_baseline.py

clotonet_code/revision_matcom/test_export_adaptive_assays.py

clotonet_code/revision_matcom/test_export_direct_matched_cost.py

clotonet_code/revision_matcom/test_manufactured.py

clotonet_code/revision_matcom/test_presentation.py

clotonet_code/v3snode.py

(84 additional files not shown) You must enabled JavaScript to view entire file list.

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.