Skip to content
All projects
ML & DataKaggle competition · University of Helsinki2025Completed

New Particle Formation Classification

5th out of 149 teams in the University of Helsinki ML course competition.

Screenshot of New Particle Formation Classification
final leaderboard rank
5th / 149
final leaderboard rank
binary accuracy
93.1%
binary accuracy
features after selection
102 → 12
features after selection

Overview

Classifying days with and without new particle formation events from 102 atmospheric measurements. Feature selection reduced the input to 12 features without losing accuracy.

Approach

  • Statistical tests, tree-based importances and L1 regularisation combined to select features.
  • Logistic regression and random forests with class weighting, because the classes were imbalanced.
  • Cross-validation for model selection. Accuracy on the harder four-class version of the task was 70.2%.
Next projectCancer Type from Mutational Signatures