Malware Clustering - Static and Dynamic Behavior Analysis
Explore innovative malware clustering techniques using static and dynamic behavior analysis, machine learning, and similarity search for robust, scalable, and repeatable results.
Hearing Without Listening: Voice Authentication with Privacy
Explore voice authentication privacy challenges and solutions, including cryptographic methods and secure binary embeddings for protecting user voice prints while enabling authentication.
How LSH Random Projection Works in Search - Python
Explore efficient similarity search using Locality Sensitive Hashing (LSH) and Random Projection. Learn how to handle large-scale data and implement approximate search techniques in Python.
Under the Hood: Key Algorithms Powering Vector Search - HNSW, LSH, and PQ
Dive into fundamental algorithms like HNSW, LSH, and PQ that power vector search and AI-driven databases, exploring their technical mechanisms and practical applications.
Reversible Transformer - GPU Memory Optimization Using ReFORMER and Reversible Residual Layers
Delve into GPU memory optimization techniques for transformers, exploring reversible residual layers and locality-sensitive hashing to enhance deep neural network efficiency and performance.
Locality Sensitive Hashing for Search with Shingling + MinHashing - Python
Learn efficient similarity search using Locality Sensitive Hashing (LSH) with shingling and MinHashing techniques in Python. Explore implementation steps, from data preprocessing to LSH tuning, for approximate nearest neighbor search.
World's Hardest Jigsaw vs. Puzzle Machine - All White
Watch a puzzle machine tackle the world's hardest jigsaw—an all-white puzzle. Explore image processing, mathematical algorithms, and innovative engineering solutions in this challenging project.
Private Nearest Neighbor Search with Sublinear Communication and Malicious Security
Explore secure and efficient private nearest neighbor search techniques, focusing on sublinear communication and protection against malicious adversaries in two-party computation scenarios.
Reformer - The Efficient Transformer
Discover how Reformer revolutionizes Transformer models, reducing memory usage and enabling processing of longer sequences through innovative techniques like Locality Sensitive Hashing and Reversible Networks.
Long-context Attention in Near-Linear Time
Explores HyperAttention, an innovative approach to tackle computational challenges in long-context LLMs, offering significant speed improvements while maintaining performance across various datasets.
Efficiently Computing Similarities to Private Datasets
Explore advanced techniques for computing similarities to private datasets, focusing on extroverted sublinear algorithms and their applications in data privacy.
O(log s)-Approximate Nearest Neighbor Search for Earth Mover's Distance
Explore advancements in nearest neighbor search for Earth Mover's Distance, improving approximation to Õ(log s) with sublinear query time using data-dependent LSH functions and tree embedding techniques.
Kernel Density Estimation through Density Constrained Near Neighbor Search
Explore kernel density estimation through near neighbor search, focusing on importance sampling, locality sensitive hashing, and data-dependent techniques for improved efficiency and accuracy.
Time Signature Based Matching for Data Fusion and Coordination Detection in Cyber Relevant Logs
Methodology for detecting automated behavior in cyber logs using temporal pattern matching, scalable through locality sensitive hashing. Applications include anomaly detection and data fusion.
Speeding Up Search With Locality Sensitive Hashing
Explore locality sensitive hashing for faster search and similarity comparisons, with practical Python implementations and real-world applications in image search and collaborative filtering.
A Quasi-Monte Carlo Algorithm for Smooth Kernel Evaluation
Explore a novel data structure for density queries, combining discrepancy theory and LSH to build sparse coresets and enhance smooth kernel evaluation algorithms.
Understanding Distance Metrics in Data Mining - Spring 2023
Explore various distance metrics in data mining, from Euclidean and Manhattan to Mahalanobis and Angular, learning their applications, limitations, and relationships to LSH algorithms.
Scaling Up AI: Embedding Physics Modeling into End-to-end Learning and Harnessing Random Projection
Explore AI-driven scientific discovery through embedding physics models in end-to-end learning and utilizing random projection for efficient simulation and gradient propagation in PDE-based systems.
Generalizations of Matrix Multiplication for Solving the Light Bulb Problem
Explore a novel approach to solving the Light Bulb Problem using tensor-based algorithms, potentially outperforming traditional matrix multiplication methods for faster data analysis and pattern recognition.