Projects#

Deep auditory encoding with self-attention to predict brain activity#

This library allows you to train a recurrent DNN (a GRU) and learn a self-attention mechanism that weighs hidden states - the resulting weighted tensor is used to predict brain activity (or whatever you choose as a target). It also contains many variations of this model type (shared attention between targets, multi-head attention etc) and some functions for visualizing the computed attention weights on a spectrogram.

The general idea is captured in this figure:

attention model

Read more in this blogpost!

A BIDS app implementing voxel-wise encoding models in fMRI using Docker#

The voxel-wise encoding BIDS app is a suite of Python tools for preprocessing fMRI and stimulus data, temporally aligning them, creating a lagged stimulus representation and training and validating voxel-wise encoding models using Ridge regression with hyperparameter search. If you don’t want to bother with a full Python installation, or want to run your analyses on a HPC cluster, you can easily use a Docker image for a smooth and reproducible workflow.

https://raw.githubusercontent.com/mjboos/voxelwiseencoding/master/scheme_BIDS_encoding.png

See this blogpost to learn more!

Jigsaw Toxic Comment Classification Challenge on Kaggle#

This was my first Kaggle challenge that I took seriously and I managed to get into the top 3% (119th place out of 4550 teams) - nothing to brag about, but a great learning experience nonetheless.

You can find the repository with all code for preprocessing the text data, training various models, doing hyperparameter search for various models, doing rigorous cross-validation and creating out-of-sample predictions that can be used for ensembling here.