# Whats the best way to classify data for Fall detector device?

I would like to know what's the best way or the function to use to classify 2 files for a fall detection device. my first file contains only TDL (Normal Movements) values and the second contains only (Fall Movements) that i took from Hexiwear ( using the Accelerator) with a python script on raspberry pi.I applied the SVM formula on both files to get one value on each 3 successive values (X,Y,Z).So far i'm using the classic classifier function classify , it generated the Knn Classifier with an accuracy of 65 % (is not really good) , but when i tried on real time with new values from the Hexiwear , its not always efficient especially with Fall movements.I know on Matlab it could be easier but since Im using the raspberry Pi , as i know , its hard to generate the Matlab code on the Pi since they are not using the same architecture.

I appreciate any suggestions or help.

Thank you all

• Wait, are you using Mathematica, or MATLAB? – J. M. will be back soon Mar 28 '18 at 15:19
• Im using Mathematica 11.3 . – Riadh Saïd Mar 28 '18 at 15:27