The code accompanying the paper entitled "Submodular Streaming in All Its Glory: Tight Approximation, Minimum Memory and Low Adaptive Complexity"
All the algorithms of the paper (single-source and multi-source settings) and required utility functions are implemented in "icml2019.py".
In "compare_streaming_algorithms.ipynb", we compare SIEVE-STREAMING++ with SIEVE-STREAMING and PREEMPTION-STREAMING, the two existing standard approaches for monotone k-cardinality submodular streaming. We evaluate the performance of these algorithms based on utility and memory complexity.
Two datasets are provided as "tweets.txt" and "YouTube_vectors.txt".
Citation: Ehsan Kazemi, Marko Mitrovic, Morteza Zadimoghaddam, Silvio Lattanzi, and Amin Karbasi. Submodular Streaming in All Its Glory: Tight Approximation, Minimum Memory and Low Adaptive Complexity. ICML, 2019.