Research Project under the mentorship of Abdiel Rivera, PhD in Electrical Engineering @Uconn
Beyond Euler: An Explainable Machine Learning Framework for Predicting and Interpreting Buckling Instabilities in Non-Ideal Materials
1 Department of Electrical and Computer Engineering, University of Connecticut, 06269, CT, USA
Correspondence: pranil.raichura@gmail.com
This repository contains the dataset and Python analysis code for the research paper, "Beyond Euler: An Explainable Machine Learning Framework for Predicting and Interpreting Buckling Instabilities in Non-Ideal Materials."
We aim to answer the question: "Can we use machine learning to accurately predict the critical buckling load of pasta columns based on physical and environmental parameters?"
This project uses an XGBoost machine learning model to predict the critical buckling load of pasta columns from geometric features. It addresses the limitations of Euler's classical buckling formula for non-ideal materials. The analysis also employs SHAP (SHapley Additive exPlanations) to interpret the model's predictions, providing insights into the underlying physics.
buckling_data.csv: The complete dataset containing 147 experimental samples of pasta buckling tests.main.py: The Python script used to perform the entire analysis./results/: The analysis results directory
git clone https://github.com/sparkyluvscode/EEI_Research_ML.git
cd EEI_Research_ML.git
pip install -r requirements.txt
python main.py