09:30 → 12:30Machine Learning lectures and tutorials¶3h
Minutes
Speaker: Adrian Bevan (QMUL)
AUK-DecisionTrees-Tutorial.pdf IntroToML_part1.pdf IntroToML_part2.pdf IntroToML_part3.pdf recording part 1 recording part 2recording part 3
We will be using the following software for tutorials: ROOT for C++ based macros (using the HEP tool TMVA), and Anaconda for all Python based work (using SciKit Learn, TensorFlow and Keras). Please see below for version information and installation instructions.
The github repository containing the examples that we will be using is also given below.
ROOT:
Please install ROOT from the ROOT Download Page. Please note that the install file is 100-200MB, and may take some time to download. If you are installing from source, the download will be faster, but you will need to ensure that you compile with appropriate options to ensure the tmva package is built.
The TMVA examples have been written for ROOT version 6.20/04 (the current PRO version).
Anaconda:
Please go to the anaconda install page and download the install file from there. The install files are several hundred MB, and so the download may take some time depending on your connection.
After installing Anaconda, please make sure you install the Spyder IDE as well as either the JupyterLab or Jupyter Notebook applications. These will be used for running the scripts and notebook examples, respectively.
After installing Anaconda, please make sure you install the following python packages:
- tensorflow
- pandas
- keras
- sklearn
- scikit-plot
- matplotlib
- numpy
- opencv
Packages can be installed from the command line using pip(3) install <package name>
, or via the anaconda navigator window by selecting "Environments", changing the pull down menu to "All" in the main window, and entering the package name under the "Search Packages" field. Having done that you should be able to click on the green button "Apply" which will create a dialogue box for the package install.
Tutorial Files:
These can be downloaded from github using the following command, or by visiting the project URL: IntroToML.
git clone https://github.com/adrianbevan/IntroToML.git
26/05/2020, 20:29 Adrian Bevan