I have some data from sensors in my phone.I have their respective battery levels at each timestamp the sensor readings were recorded in phone.My aim is to be able to predict, lets say that i have sensors readings for 12 hours,what will be my battery level after a day? (I have training data,which indicated user patterns,i am also given a time period to predict the users battery level in the future).
I can create some kind of model,to train my mdoel to predict the current battery level using training data(clustering may be?)
I do not know how to predict the future,I mean theoretically how would it work?
IN addition,how can I use the pattern given to me ,and the model created together to predict the battery level in the future? I was thinking of matching the patterns, to the regression model i will create on the timestamp, and then look at the time difference,add it and find out battery reading acc to my model,but Im not sure how accurate this would be.
Any help will be appreciated.