Skip to content

Using machine learning to predict soil properties via infrared spec

Notifications You must be signed in to change notification settings

Michael-Bar/Machine_learning_infrared

Repository files navigation

Overview

These scripts use mid-infrared spectroscopy data to predict soil properties from a library of wet chemistry results.

The scripts run a PLS and RF algorithm on spectra, the code is augmented with Monte Carlo error estimation and a generalisable code base (to allow more models to be tested)

About

Using machine learning to predict soil properties via infrared spec

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages