High-Energy Particle Signal Classification

Kisel, I., Lakos, R., & Zischka, G. (2025). Deep-Learning-Based Optimization of the Signal/Background Ratio for Λ Particles in the CBM Experiment at FAIR. Algorithms, 18(4). https://doi.org/10.3390/a18040229

• Developed and trained 20+ classification models using different architectures of Logistic Regression, SVM, XGBoost, and Neural Networks (TensorFlow, sklearn MLPClassifiers) to identify Higgs boson signals from background processes.
• Cleaned and scaled a dataset of more than 600,000 entries, each represented by 28 features, engineering high-level physical descriptors from raw kinematic measurements.
• Analyzed overfitting and underfitting across algorithms by studying training/testing/cross-validation accuracy and F1 scores while tuning parameters, visualizing findings through comparative graphs.
• Achieved a best-performing model with 76.74% training accuracy and 75.52% testing accuracy, documented in a technical report and presented to faculty.