This is the work that was done as a part of Major Project when I was a B.Tech student at NITK Surathkal in 2014. This was for physionet 2014 challenge http://physionet.org/challenge/2014/ Heart beat detection in Multimodal data using Deep learning(simple stuff but was cool) and XGBoost(didn't really beat the performance of simple logistic regression). The results are really good and resulted in 1st place in all 3 phases of the competion. The project is in octave and python. The python version uses XGBoost and Octave version uses Logistic regression from Deep learning toolbox.
To run this you need to download data ECG or BP data from physionet-- links are available on competion web page - http://physionet.org/challenge/2014/
Some of the paths mentioned in the code are local, please make chages accordingly if you find an error.