Side projects

NayesDog: RSS reader with Naive Bayes powered recommendations

Have You Ever Felt Like You Were Drowning in an endless sea of data while you try to retrieve valuable information? NayesDog is a recent project developed with Ilya Prokin that aims to provide a smart and helpful RSS reader. As for most RSS readers, users can subscribe to their favorite feeds by entering the feed’s URI and then the reader inspects and shows the chosen feeds when required. However, unlike most readers, NayesDog reads the entries and sorts them according to your preferences before showing them. Therefore users may focus on news that have been classified as more interesting for them by the algorithm. NayesDog first converts feed entries into bag of words and then uses a Naive Bayes classifier to correlate words with the user preferences. Indeed, preference buttons (like / dislike) allows users to give a feedback to the algorithm regarding their preferences. The algorithm has been designed to adapt to evolvable preferences of users.

Time Series Analysis

TimeSeriesAnalysis is recent project created with Brice Denoun. The aim of this project is to extract knowledge from EEG signals captured by low-cost sensors using deep learning, machine learning and time series analysis in order to prevent for example seizures and cerebrovascular accident dynamically. Our purpose is to make the software as useful, light and portable as possible so it could produce suitable results while being executed in a small device (e.g., a raspberry pi).