New Accepted Solution: Top Quark Detection with Deep CNN

[robertogl]

🎉 New Accepted Solution: Top Quark Detection with Deep CNN

Submitted by from .

Overview. This project addresses the challenge of identifying top quark events within large-scale particle collision datasets using deep learning techniques. Leveraging MATLAB’s Deep Learning Toolbox, the student developed a Convolutional Neural Network (CNN) capable of distinguishing top quark jets from background noise, enabling more accurate tagging and improved event classification performance.

Important

Why it won: Innovative adaptation of CNN architectures for high-energy physics data, achieving significant performance gains in quark tagging.

Tip

Standout results: The student utilized transfer learning with pretrained CNN layers and extensive data preprocessing to manage large, noisy datasets, achieving strong accuracy and generalization.

At a glance

• Dataset/Models: Simulated top quark collision datasets from CERN Open Data
• Toolboxes: Deep Learning Toolbox, Parallel Computing Toolbox
• Training/Eval: CNN model with data augmentation and batch normalization
• Deployment: Optimized inference for high-throughput tagging

Accepted to the MATLAB & Simulink Challenge Project Hub. Congratulations to from !