US 11,720,565 B2
Automated query predicate selectivity prediction using machine learning models
Mohamad F. Kalil, Ottawa (CA); Calisto Zuzarte, Pickering (CA); Mustafa Dawoud, London (CA); Mohammed Fahd Alhamid, North York (CA); Vincent Corvinelli, Mississauga (CA); Wai Keat Tan, Vancouver (CA); and Ronghao Yang, Toronto (CA)
Assigned to International Business Machines Corporation, Armonk, NY (US)
Filed by INTERNATIONAL BUSINESS MACHINES CORPORATION, Armonk, NY (US)
Filed on Aug. 27, 2020, as Appl. No. 17/4,225.
Prior Publication US 2022/0067045 A1, Mar. 3, 2022
Int. Cl. G06F 16/2453 (2019.01)
CPC G06F 16/24545 (2019.01) [G06F 16/24537 (2019.01); G06F 16/24544 (2019.01)] 26 Claims
OG exemplary drawing
 
1. A method for cardinality estimation using machine learning, the method comprising:
accessing database relations;
collecting a random sample from each of the database relations;
generating training data for creating a cumulative frequency function (CFF) model from the random sample collected from each of the database relations, the cumulative frequency function (CFF) model predicted using a machine learning model;
creating the cumulative frequency function (CFF) model from the generated training data using the machine learning model; and
predicting a cardinality estimation of a plurality of structured query language (SQL) predicates, the cardinality estimation based upon the cumulative frequency function (CFF) model.