A sample SVM library for Android. https://github.com/yctung/AndroidLibSVM
- Train a model by picking a user selected training file from the data folder (containing both walking and stationary) to build a model using only acceleration. The format of the file as follows. 0th row {first name, last name, mobile, email, gender, age} 1st row {timestamp, lat, long, accelx, accely, accelz, label} 2nd row onwards contain values as per the format in the 1st row
- Read a user selected test file from the data folder containing the same information and apply the learned model for prediction.
- Output a file in the data folder containing the prediction in the following format 0th row {first name, last name, mobile, email, gender, age} 1st row {timestamp, lat, long, accelx, accely, accelz, label, prediction} 2nd row onwards contain values as per the format in the 1st row