User Guide

A comprehensive guide

This guide explains how to get a functionnal neural network, from data opening to analyzing results.

• Reading data
• Initializing HIBOU
• Filtering the data
• Splitting the data and train the ANN
• Analyze output

Working with gradient boosted trees

HIBOU also provides a gradient boosted trees (GBT) model, the manipulations are very similar, but some differences are important. Please note that the GBT can only estimate one flux at a time, so you’ll need two simulations to estimate H and LE.

• Train the GBT