Using decision trees and their ensembles for analysis of NIR spectroscopic data

Advanced machine learning methods, like convolutional neural networks and decision trees, became extremely popular in the last decade. This, first of all, is directly related to the current boom in Big data analysis, where traditional statistical methods are not efficient. According to the kaggle.com — the most popular online resource for Big data problems and solutions — methods based on decision trees and their ensembles are most widely used for solving the problems.

It can be noted that the decision trees and convolutional neural networks are not very popular in Chemometrics. One of the reasons for that is the landscape of the data matrix: the modern machine learning methods need number of measurements much larger than the number of variables to avoid overfitting, which is opposite to the layout of the data we usually deal with. Another drawback is a lack of interactive instruments for exploring and interpretation of the models.

In this presentation, we are going to discuss an applicability of decision trees based methods (including gradient boosting) for solving classification and regression tasks with NIR spectra as predictors. We will cover such aspects as evaluation, optimization and validation of models, sensitivity to outliers and selection of most important variables.