theolepage/ember-classification

Collection of machine learning methods to classify the EMBER dataset.

ESLR Recruitment Project 2019

Resources

Subject: https://drive.google.com/open?id=1VV6E7x1FfE8PHIY0-B6TA3uWLK6knjhv
Dataset: https://drive.google.com/open?id=1e6X56yLlmXzaeavJ-Nn8oG2kEtCHnC4T

k-means

Usage

You have to put the ember dataset in ember/.

To run kmeans on the dataset and check its accuracy, follow the steps below.

cd kmeans/
make test

To run kmeans only, type the folowing command: ./kmeans 3 20 1.0 2351 900000 ../ember/Xtrain.dat out.dat.

Classifiers and neural networks

Usage

Jupyter notebooks and .py source files are located in specific subfolders. However, we suggest using Google Colab to run our experiments.

Step 2: Classification with machine learning

• Path: python/step2-classifier_machine_learning
• Google Colab: https://colab.research.google.com/drive/1EAA0eecyZ4oqouvi_dwydWBAa3sEV1uI

Step 3: Classification using Deep Neural Network

• Path: python/step3-deep_neural_network
• Google Colab: https://colab.research.google.com/drive/1TwfyeQhqQqtxPH224KzFxeUSDagQE9Mp

Bonus

Usage

Jupyter notebooks and .py source files are located in specific subfolders. However, we suggest using Google Colab to run our experiments.

Metric Learning: Siamese network

• Path: python/bonus_siamese_network
• Google Colab: https://colab.research.google.com/drive/1G2sZT-5-1WpdbVyL_Ic0qANF4iSpyg2F