The Higgs boson machine-learning challenge aims to explore the potential of advanced machine-learning methods to improve the analysis of data produced by the experiment. This project relies on high dimensional data and targets the creation of a binary classifier that classifies events as an observation of the decay of the Higgs particle into two tau particles.
- Asli Andréa : andrea.asli@epfl.ch
- Berquet Romain : romain.berquet@epfl.ch
- Chammas Michel : michel.chammas@epfl.ch
-
project1_description.pdf : Project guidelines.
-
project_report : Written report highlighting the most important findings obtained.
-
data : provides two files in the .csv format : test.csv (test set) and train.csv (training set).
-
scripts : Provides all the scripts that are needed to implement the project's methods.
-
run.py : Runs the algorithms and provides the predictions in the output folder.
-
implementations.py : Provides all required machine learning methods.
-
proj1_helpers.py : Provides the helper methods used by the code : loading the data, loss and gradient computation....
-
data_processing.py : Provides methods for preprocessing the dataset before using any algorithm on it.
-
optimisation.py : Provides the optimal degree and lambda using k-fold cross-validation (10-fold).
-
validation.py : Provides methods used to execute cross-validation (Data split).
-
parameters.py : Contains the parameters we used.
-