What kind of particle was that?
Using Machine Learning to identify particles.

Let's imagine that some particles are going into a detector.
Each particle breaks up into many fragments, and the detector counts the number of fragments, and measures how evenly the energy is shared between the fragments.
You turn the detector on, and there seem to be 2 kinds of particles. Can you train a neural network (a machine learning tool) to separate the particles for you?

Epoch

Observations

Change universe! What is the behavior of the particles you are trying to seperate?

Features

Let the network use number of fragments, energy sharing, or both?

Click anywhere to edit.
Weight/Bias is 0.2.
This is the output from one neuron. Hover to see it larger.
The outputs are mixed with varying weights, shown by the thickness of the lines.

Output

Test loss
Training loss
Point color indicates the true identity of the particle, backgound color indicates the current prediction of the particle's identity.

Quick instructions;

Click on things!
Things you can click on to change the challange we are trying to solve are;

Things you can click on to adapt the machine learning tool, a "feed forward neural network", are; Things you can click on to train and evaluate the neural network are;

Credits

Heavly cribbed from the Tensorflow Playground by Google, which was created by Daniel Smilkov and Shan Carter. Adapted for the DESY open day by Henry Day-Hall.