Published October 5, 2026 | Version v2.2.0

kanfood: Kolmogorov–Arnold networks as closed-form calibrations for food spectroscopy

  • 1. Yildiz Technical University, Department of Mechatronics Engineering
  • 2. Yildiz Technical University, Department of Food Engineering

Description

Code, results and the open tahini ATR-FTIR adulteration dataset for a leakage-free comparison of a Kolmogorov-Arnold network with PLS, SVR, random forest, MLP and 1-D CNN calibrations (tahini by ATR-FTIR, mango dry matter by NIR), including the extraction of closed-form calibration equations.

Notes

If you use this software or the tahini dataset, please cite both the article and this repository.

Files

abdulhamidbatayhi123/kanfood-tahini-mango--v2.2.0.zip

Files (32.2 MB)

Name Size
md5:a8bcf11ed25f5c9aa720fa3071c0e611
32.2 MB Preview Download

Additional details