Welcome!

This tool automates the process of applying machine learning to geochemical data. You can use it to discriminate data by classes (e.g. rock classification) or to create clusters (e.g. analysis of geochemical signatures of different locations). We will be guiding you through the process!

This tool is based on the following paper (open access):

Alférez, G.H., Esteban, O.A., Clausen, B.L., & Martínez Ardila, A.M. (2022). Automated machine learning pipeline for geochemical analysis. Earth Science Informatics. https://doi.org/10.1007/s12145-022-00821-8

First, let´s start by uploading the dataset. The data must meet the following requirements:

• Be in CSV format
• Only numeric data features (oxides must be entered as percentages and single elements must be entered as parts per million - ppm)
• Only string data values for classes (for classification only)
• Every row must belong to a class (for classification only)
• Only named variables in the first row

As reference, download & review sample input data:

• Cinco.csv
• CincoTrain.csv

Upload