Weighted Random Sampling over Data Streams
- WRS algorithms for data streams
- base class: StreamSampler class
- Algorithm A-Chao: StreamSamplerChao
- Algorithm A-Chao with Jumps: StreamSamplerChaoWithJumps
- Algorithm A-ES: StreamSamplerES
- Algorithm A-ES with Exponential Jumps: StreamSamplerESWithJumps
- Preliminary Implementation of the Algorithm in Java, and
- Execution Examples
- Downloads:
- The StreamSampler classes with the algorithms in Java (Eclipse project archive)
- A demo project using the StreamSampler classes (Eclipse demo project archive)
- A related report: P.S Efraimidis. Weigthed Random Sampling over Data Streams, http://arxiv.org/abs/1012.0256.
The original project:
Weighted Random Sampling (WRS) with a Reservoir
- WRS Algorithms
- Efficient Weighted Random Sampling
- with one-pass over unknown populations (for example data streams)
- high pararellizable
- Preliminary Implementation of the Algorithm in Java, and
- Execution Examples
- Efficient Weighted Random Sampling
- Download the application code (WinZip Archive)
- A related paper: P.S Efraimidis and P. Spirakis. Weigthed Random Sampling with a Reservoir, Information Processing Letters 97 (5), pp 181-195, 2005.
