Machine Learning Tutorial: ATLAS UK Collaboration Meeting, January 2020: This page is maintained by Dr Adrian Bevan
This web page contains a set of slides and a tar file of code examples (ROOT macros and
Jupyter notebooks) used for the 2020 ATLAS UK Collaboration Meeting tutorial on machine learning.
If you are interested in my other machine learning lectures please see the corresponding section of my
teaching page.
github
The code examples and data for this tutorial are on git and can be downloaded via https://github.com/adrianbevan/ATLAS-UK-ML. Alternatively use:
git clone https://github.com/adrianbevan/ATLAS-UK-ML.git
Tar File and file content
The code examples for this tutorial can be obtained by downloading the file ATLAS-UK.tar and unpacking this using the command:
tar -xvf ATLAS-UK.tar
This will unpack three directories and a README file. The directories are data, NN and BDT and they contain the following:
- data/train_sml_bg.csv -- background training data
- data/train_sml_sig.csv -- signal training data
- BDT/TMVAClassification.C -- this file is the main BDT training macro
- BDT/TMVAClassificationApplication.C -- this file applies the trained BDT to data
- BDT/Compare.C -- this file compares 4 different types of BDT training with the same data
- BDT/Htautau.C -- this file is for use with the Kaggle Higgs Data Challenge data
- NN/LinearRegression.ipynb -- this is a linear regression minimisation problem
- NN/FunctionApproximation.ipynb -- this is a simple neural network with one input, 50 nodes in a single layer and one output.
- NN/FunctionApproximation2.ipynb -- this is a Multi Layer Perceptron example (added after the tutorial)
Boosted Decision Trees (BDTs)
- Decision Trees (Background) : (pdf)
- Decision Trees (Tutorial) : (pdf)
Neural Networks (NNs)
- Neural Networks (Background) : (pdf)
- Neural Networks (Tutorial) : (pdf)
Back to the home-page