Data Mining: Choosing k for MinHashing - Spring 2023

Data Mining: Choosing k for MinHashing - Spring 2023

Learn advanced data mining concepts in this 21-minute lecture focusing on Min Hashing techniques and statistical foundations. Explore the mathematical principles behind choosing optimal k values for Min Hashing, including Probably Approximately Correct (PAC) learning, Central Limit Theorem, and Chernoff-Hoeffding Inequality. Master the application of these theoretical concepts to obtain accurate Jaccard Similarity estimates through Min Hashing. Delve into practical implementations while understanding the statistical guarantees that make Min Hashing a powerful technique in data mining applications.