About this question

Count-Min Sketch for Per-Symbol Volume Estimation

Hard · data_structures · Quant Developer interview question · count-min-sketch, probabilistic, frequency-estimation, data_structures, hashing, volume

A Count-Min Sketch is a probabilistic data structure used to estimate the frequency of events in a data stream with sub-linear space. In high-frequency trading, it is used for real-time surveillance tasks like tracking per-symbol trade volumes where memory is a critical constraint. The sketch provides an unbiased overestimate of the true frequency, never undercounting but with a controlled probability of overcounting due to hash collisions. Two Sigma uses sketches for high-throughput market-data