Code 30 lines. Get 100 USD
K-Means Clustering
App Name:
[login to view URL]
App Description:
Use k-means to group 400 addresses into 5 groups.
1. App starts and plots the 400 addresses
2. User inputs 5 starting cluster points
3. K-cluster algorithm runs and groups addresses into appropriate clusters
4. 5 groups (80 addresses each) are generated
App Requirements:
1. Distance calculated using Google Maps walking/driving distance; NOT, geometric distance.
(refer to attached picture)
A and B are two locations on a map. Blue line indicates to the walking distance between points A and B calculated using Google Maps API. Red line indicates the geometric distance between point A and B. For this project, K-Means will use the Google Maps API (blue line) distance to determine appropriate clustering, NOT the geometric distance (red line).
2. Developed for Android Platform (2.3)
3. Must use/load Weka Library (unless you can find something better)
4. Cluster minimum 400 addresses in Google Maps into 5 clusters of 80 addresses. Sample address list will be provided.
4. Initial centroid will be generated by touch input on the map.
5. Number (#) of clusters must be dynamic (can be changed to x # of clusters)
6.. When you respond to this posting, please type in the code
BR3P2?k0xdr
or we will consider your response spam and discard it.
7. Though we expect a very short program, we don't care if your program is either under or over 30 lines. If it works, you get the money.
8. All code must be posted to a CVS repository that we will provide.
![][1]
Important Note:
- If you are hired, we will have consistent year-long work for you.
- We also do not mind for you to work on other projects simultaneously.