I have a set of binary vectors where each vector represents one day of occupancy in a house and consists of 48 elements (each element for 30 minutes of the day). Each element can be 1 meaning that house was occupied and 0 for non occupied house.

My task is to predict the next day based on the history of the same days (Monday from history of Mondays etc.). So far I am using hamming distance to find 5 most similar days in the history and from them I calculate the probabilities of the occupancy as a mean of those 5 numbers. When the probability is higher than some X, in my case 0.4, I predict it to be occupied.

But there is definitely some more efficient way to do this, any algorithms that would capture the trend in the history?

  • $\begingroup$ What is the underlying model/assumption? $\endgroup$ – Raphael Apr 30 '13 at 22:01
  • $\begingroup$ How much data do you have? $\endgroup$ – alto May 1 '13 at 0:29
  • $\begingroup$ enough data- 3 months of data with around 80 events per day $\endgroup$ – totpiko May 1 '13 at 13:54

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.