I am trying to find an approach to filter out the jitter points in GPS data points that are collected from an Android application.
I have been researching for a while and all of the solutions that I have found so far assume dealing with dense data which is the extreme opposite at my case. There are some extreme cases that are so obvious as the cases for infeasible velocity. However, there are many others that can not be detected trivially especially when the data is sparse.
I have all the GPS points in a pandas dataframe where for each GPS data point, I have its time stamp, latitude, longitude, distance and speed from its previous and preceding points. Also, for the whole daataframe, I can detect stops, trips and common places for the user.
Any ideas would be appreciated especially if it does not include any back end call.
