Real-time cardinality estimator built for the Future of Database Programming Contest. Implements histogram-based methods, HyperLogLog, MCV tracking, and optimized data structures for high-performance. Achieved first place in the competition

Triangle Counting (DataFoutain, 三角形图计算算法设计及性能优化, top 11)

Contains the code and examples for my article on Medium, which explains how to handle data skew in Apache Spark to improve performance.

In Apache Storm, when sources are skewed, the PKG is not enough to guarantee the load balancing between down-stream bolts. Thus, we advance intellectual novelties to provide load balancing at global scale in Information Flow Processing Systems called "Back Propagation and Periodical Calibration".