Raspberry Pi interest group



The Raspberry Pi has emerged as a cheap and very attractive platform for learning and teaching in Computer Science for kids, beginners and everybody else with particular need whatever it may be.

We, at University of Rouen, noticed that the Raspberry Pi community has this this outstanding capacity of innovation and experience in electronics. However, it seems that there lacks capacities of intelligent data processing, mining, classification, detection, recognition or other Artificial Intelligence like functionalities. Data analysis tools for statistics, Pattern Recognition, Computer Vision and Machine Learning have been widely developed the last decades but mostly understood and used by professionals.

Our aim is to deliver the best of these fields to the Raspberry Pi community in form of code and modules, tutorials and simplified technical articles. All with the goal to be used, re-used on Raspberry Pi or any other low-power device. We believe that this is a win-win deal both for Machine Learning community and outer world currently composed of enthusiasts in electronics.

We open this box of treasures for wide public use through the Raspberry Pi!


We have working experience in Pattern Recognition and Classification, Statistical Learning from data and Machine Learning in general. We have been exposed to the world of research for years that allows us to easily browse research publications, algorithms and see their underlying workings and statistical underpinnings.

Some of our fields of expertise:

  1. Linear and non-linear models for classification and regression;

  2. Kernel methods (Support Vector Machines)

  3. Trees and Ensemble methods (Random Forests)

  4. Unsupervised learning from data (clustering, linear and non-linear dimensionality reduction)

  5. Probabilistic approaches (Belief Networks, Graphical models, HMM, Mixture models)

  6. Learning from high dimensionality data


We are open and attentive for any intelligent data processing needs within the Raspberry Pi community and beyond. The contributions are currently purely experimental and guided by our personal interest. The code, demos and other materials may be shared through our blogs or by any other mean.

Feel free to contact us whenever it concerns data mining, learning from data, decision making on data and similar.

The Raspberry Pi Interest Group of University of Rouen, France


October-December 2012:

Developed Extreme Learning Machine [WWW] classifier for Raspberry Pi. Provides memory and computation power saving classifier training and prediction capabilities. [GITHUB]

February-April 2013:

Evaluated and adapted various sensors (accelerometers, gyroscopes etc), display modules (LED, LCD) and developed portable prototype for wearable data recording (backpack version).