Tunnel-Boring Machine Rate of Penetration Predictions via Machine Learning

21 machine learning models for predicting tunnel-boring machine rate of penetration, achieving R² values up to 0.97. The research by Arsham Moayedi Far, last updated in May 2026, applies feature importance analysis to streamline geology-related input parameters. Models were tuned for generalization and validated on out-of-sample datasets.

Use Cases

• Predicting tunnel-boring machine performance based on geological parameters mentioned in the description
• Comparing feature importance methods like decision trees and permutation importance for geotechnical data
• Building ensemble models for rate-of-penetration estimation using the described voting regressor approach
• Tuning hyperparameters for geotechnical ML models to improve generalization

Strengths

• Includes 21 generated predictive models with reported high performance (R² up to 0.97)
• Workflow specifically designed to reduce overfitting through hyperparameter tuning and cross-validation
• Models were tested on out-of-sample and unseen datasets for validation

Limitations

• Column-level documentation is absent; field semantics must be inferred after download
• Row count is unknown, which may limit suitability assessment
• The dataset is small in size (328.6 KB), indicating a limited scope of data

Provenance

Source

Arsham Moayedi Far via figshare

Collection Method

Research study applying feature importance and machine learning algorithms to geotechnical data.

Freshness

Last updated 2026-05-09 07:49:03