- Signal Processing
-
Going through fundamental tools of signal processing using a singal from a smartphone's accelerometer and gyroscope. Basic statistical properties of the signal such as whether it is stationary or its ergodicity are checked. Next, multivariate statistics such as the (self) correlation is computed and plotted over several shifts. Finally, several convolutions are applied in the time and frequency domain. A a simple square low pass filter, e.g. sharply cutting off higher frequencies in the frequency domain, is applied and compared to a second order Butterworth filter which doesn't show the same ringing in the transformed signal.
-
Predictive Modelling
-
Deep Learning