multi-learn / summit

Supervised MultiModal Integration Tool's Readme

This project aims to be an easy-to-use solution to run a prior benchmark on a dataset and evaluate mono- & multi-view algorithms capacity to classify it correctly.

Getting Started

SuMMIT has been designed and uses continuous integration for Linux platforms (ubuntu 18.04), but we try to keep it as compatible as possible with Mac and Windows.

Platform	Last positive test
Linux
Mac	1st of May, 2020
Windows	1st of May, 2020

Prerequisites

To be able to use this project, you'll need :

• Python 3

And the following python modules will be automatically installed :

• numpy, scipy,
• matplotlib - Used to plot results,
• sklearn - Used for the monoview classifiers,
• joblib - Used to compute on multiple threads,
• h5py - Used to generate HDF5 datasets on hard drive and use them to spare RAM,
• pickle - Used to store some results,
• pandas - Used to manipulate data efficiently,
• six -
• m2r - Used to generate documentation from the readme,
• docutils - Used to generate documentation,
• pyyaml - Used to read the config files,
• plotly - Used to generate interactive HTML visuals,
• tabulate - Used to generated the confusion matrix.
• pyscm-ml -

Installing

Once you cloned the project from the gitlab repository, you just have to use :

cd path/to/summit/
pip install -e .

In the summit directory to install SuMMIT and its dependencies.

Running the tests

To run the test suite of SuMMIT, run :

cd path/to/summit
pip install -e .[dev]
pytest

The coverage report is automatically generated and stored in the htmlcov/ directory

Building the documentation

To locally build the documentation run :

cd path/to/summit
pip install -e .[doc]
python setup.py build_sphinx

The built html files will be stored in path/to/summit/build/sphinx/html

Running on simulated data

For your first go with SuMMIT, you can run it on simulated data with

python
>>> from summit.execute import execute
>>> execute("example 1")

This will run the benchmark of documentation's Example 1.

For more information about the examples, see the documentation. Results will, by default, be stored in the results directory of the installation path : path/to/summit/multiview_platform/examples/results.

The documentation proposes a detailed interpretation of the results and arguments of SuMMIT through 6 tutorials.

Dataset compatibility

In order to start a benchmark on your own dataset, you need to format it so SuMMIT can use it. To do so, a python script is provided.

For more information, see Example 5

Running on your dataset

Once you have formatted your dataset, to run SuMMIT on it you need to modify the config file as

name: ["your_file_name"]
pathf: "path/to/your/dataset"

It is however highly recommended to follow the documentation's tutorials to learn the use of each parameter.

Authors

• Baptiste BAUVIN
• Dominique BENIELLI
• Alexis PROD'HOMME

summit's Issues

Would you mind providing your experiments code

Hello, I am very interested to the Part 3.4. "Real World Use Case : Multi-Omic Study" in you article, where you used summit to solve multi-view imbalanced problems. But I don't exactly know how you achieve it in code. Particularly, I don't know how to add imbalanced_bagging and $\mu CoMBo$ into the model. So could you give me your experiment code. It will help me a lot.