Destination prediction based on past movement history
In this paper we explore destination prediction using the past movement history, where the history is built using machine learning. Matching of the history with the user's movement is done through a simple pattern recognition technique. The system includes an algorithm that simulates incoming d...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Published: |
2020
|
| Subjects: | |
| Online Access: | http://eprints.sunway.edu.my/2399/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | In this paper we explore destination prediction using the past movement history, where the history is built using machine learning. Matching of the history with the user's movement is done through a simple pattern recognition technique. The system includes an algorithm that simulates incoming data from a user and matches on a window by window basis to the history. After matching it attempts to narrow down the possible final destination by eliminating paths from history that do not meet a set threshold for the next round matching. The experiment was carried out in Sunway University building on two separate floors. The results are promising and we also look at the combination of variables to use for such a use case. |
|---|