Advances in Data Science : Women in Data Science and Mathematics (WiSDM) 2023

著者名：Garcia-Cardona, Cristina (EDT)/Lee, Harlin (EDT)

Springer（2025/09/26発売）

• 言語：ENG
• ISBN：9783031878039
• eISBN：9783031878046

Description

This volume features recent advances in data science ranging from algebraic geometry used for existence and uniqueness proofs of low rank approximations for tensor data, to category theory used for natural language processing applications, to approximation and optimization frameworks developed for convergence and robustness guarantees for deep neural networks. It provides ideas, methods, and tools developed in inherently interdisciplinary research problems requiring mathematics, computer science and data domain expertise. It also presents original results tackling real-world problems with immediate applications in industry and government.

Contributions are based on the third Women in Data Science and Mathematics (WiSDM) Research collaboration Workshop that took place between August 7 and August 11, 2023 at the Institute for Pure & Applied Mathematics (IPAM) in Los Angeles, California, US. The submissions from the workshop and related groups constitute a valuable source for readers who are interested in mathematically-founded approaches to modeling data for exploration, understanding and prediction.

Table of Contents

Chapter 1: Randomized Iterative Methods for Tensor Regression Under the t-product.- Chapter 2: Matrix exponentials: Lie-Trotter-Suzuki fractal decomposition, Gauss Runge-Kutta polynomial formulation, and compressible features.- Chapter 3: An exploration of graph distances, graph curvature, and applications to network analysis.- Chapter 4: Time-Varying Graph Signal Recovery Using High-Order Smoothness and Adaptive Low-rankness.- Chapter 5: Graph-Directed Topic Models of Text Documents.- Chapter 6: Linear independent component analysis in Wasserstein space.- Chapter 7: Faster Hodgerank Approximation Algorithm for Statistical Ranking and User Recommendation Problems.- Chapter 8: A Comparison Study of Graph Laplacian Computation.- Chapter 9: Supervised Dimension Reduction via Local Gradient Elongation.- Chapter 10: Reducing NLP Model Embeddings for Deployment in Embedded Systems.- Chapter 11: Automated extraction of roadside slope from aerial LiDAR data in rural North Carolina.- Chapter 12: A non-parametric optimal design algorithm for population pharmacokinetics.- Chapter 13: Unrolling Deep Learning End-to-End Method for Phase Retrieval.- Chapter 14: Performance Analysis of MFCC and wav2vec on Stuttering Data.- Chapter 15: Active Learning for Reducing Gender Gaps in Undergraduate Computing and Data Science.- Chapter 16: Quantifying and Documenting Gender-Based Inequalities in the Mathematical Sciences in the United States.