Big Streaming Data Sampling and Optimization

This research addresses and resolves the issues with the confidence level of sampled big streaming data that is dynamic with respect to the speed of the streaming data and the dynamically changing sample space. Based on a preliminary work and results from [ 8 ], this research focuses more on the confidence level and threshold of dynamic size of the population in order to ensure a better confidence level of the sampled data with respect to a few variables such as speed of the streaming data, population size dynamic over time, sample space (or size), speed of sampling algorithm, size of streaming data, and time duration of data streaming. Theoretical thresholds of the processing of big streaming data with respect to a set of variables as mentioned above are identified in an effort for optimization. Simulation results along with experimental results are provided to validate the efficacy of the proposed theoretical thresholds.