Classification of the data using a convolutional neural network

The aim of the workshop is to introduce students into the application of machine learning methods to classify EUSO-SPB1 data, but in this case, by using a Convolutional neural network. The presentation includes a brief introduction to the topic and examples of expected results. In this exercise, students can experiment with neural network architectures and afterward compare their results.

The Air Shower Pattern Recognition summer school consists of the following five workshops:

1. Introduction to the JEM-EUSO and running air shower event simulations
2. Reading the detector data
3. Implementation of a feature extraction method
4. Classification of the data through the extracted features
5. Classification of the data using a convolutional neural network

Slides

Practical activity

1. Creating CNN in Keras 
2. Training of the network
3. Comparison a the decision tree-based classifier