trans_about.txt for WISDM_Act_v1.1 dataset See readme.txt for information about the WISDM Lab, rights, and other general information. For our transformation process, we take 10 seconds worth of accelerometer samples (200 records/lines in the raw file) and transform them into a single example/tuple of 46 values. Most of the features we generate are simple statistical measures. Associated tasks: classification * Number of examples: 5,424 * Number of attributes: 46 * Missing attribute values: None * Class distribution: * Walking -> 2,082 -> 38.4%, * Jogging -> 1,626 -> 30.0%, * Upstairs -> 633 -> 11.7%, * Downstairs -> 529 -> 9.8%, * Sitting -> 307 -> 5.7%, * Standing -> 247 -> 4.6% * transformed.arff follows the Attribute-Relation File Format specified [here](http://weka.wikispaces.com/ARFF+%28stable+version%29) * Field descriptions: To see the field definitions, read the arff file's header. * UNIQUE_ID: just that, a unique identifier for each tuple. We exclude this field when making predictions * user is the id number of the user that the data is from. * X0..x9, Y0..Y9, Z0..Z9 are bins, their values are the fraction of accelerometer samples that fell within that bin * XAVG, YAVG, ZAVG are the average x, y, and z values over the 200 records in the example. * XPEAK, YPEAK, ZPEAK are approximations of the dominant frequency. First, the greatest value in the series is identified, then all local peak values within 10% of its amplitude are identified. If the number of peaks is less than 3, then the threshhold is lowered until at least 3 peaks can be found. The times between consecutive peaks are summed and divided by the number of peaks. * XABSOLDEV, YABSOLDEV, ZABSOLDEV are the average absolute deviations from the mean value for each axis. * XSTANDDEV, YSTANDDEV, ZSTANDDEV are the standard deviations for each axis. * RESULTANT is the average of the square roots of the sum of the values of each axis squared √(xi^2 + yi^2 + zi^2). * class is the activity that the user was performing during this example. For a detailed specification, see section 2.2 of: Jennifer R. Kwapisz, Gary M. Weiss and Samuel A. Moore (2010). ["Activity Recognition using Cell Phone Accelerometers"](http://www.cis.fordham.edu/wisdm/public_files/sensorKDD-2010.pdf) Proceedings of the Fourth International Workshop on Knowledge Discovery from Sensor Data (at KDD-10), Washington DC.