Classical Learning

Classical learning theory seeks to understand and improve the generalization ability and robustness of machine learning models. Current research focuses on addressing limitations of traditional assumptions by analyzing generalization error in various contexts, including sparse mixture-of-experts models and the impact of data properties like smoothness and sparsity. This work aims to develop more reliable and efficient learning algorithms by investigating factors such as model complexity (e.g., boundary piece count), parameterization choices, and the influence of data distribution on learning performance, ultimately leading to more robust and accurate machine learning systems.

Papers

August 28, 2024

Fairness, Accuracy, and Unreliable Data
Kevin Stangl
Machine Learning Procedural Fairness Empirical Risk Minimization Robust Algorithm Unreliable Source Strategic Classification Classical Learning

June 9, 2024

General Distribution Learning: A theoretical framework for Deep Learning
Binchuan Qi
Deep Learning Machine Learning Theoretical Framework Classical Learning GD Learning

May 12, 2024

Separable Power of Classical and Quantum Learning Protocols Through the Lens of No-Free-Lunch Theorem
Xinbiao Wang, Yuxuan Du, Kecheng Liu, Yong Luo, Bo Du, Dacheng Tao
Camera Lens Quantum Machine Learning Model Quantum Learning Separable Data Classical Learning

March 26, 2024

Generalization Error Analysis for Sparse Mixture-of-Experts: A Preliminary Study
Jinze Zhao, Peihao Wang, Zhangyang Wang
Many Sparse Sparsity Increase Generalization Error Preliminary Study Sparse Mixture of Expert Expert Selection Classical Learning

March 5, 2024

Proper vs Improper Quantum PAC learning
Ashwin Nayak, Pulkit Sinha
Sample Complexity Quantum Learning Classical Learning Improper Learning

December 31, 2023

On the Necessity of Metalearning: Learning Suitable Parameterizations for Learning Processes
Massinissa Hamidi, Aomar Osmani
Inductive Bias Community Need Learning Process Classical Learning Metalearning Algorithm

June 15, 2023

Exact Count of Boundary Pieces of ReLU Classifiers: Towards the Proper Complexity Measure for Classification
Paweł Piwek, Adam Klukowski, Tianyang Hu
Native Robustness Classification Code Decision Boundary Robust Training Borda Counting ReLU Neural Network Complexity Measure Non Contiguous Piece Classical Learning

May 25, 2023

How many samples are needed to leverage smoothness?
Vivien Cabannes, Stefano Vigogna
Generalization Error Statistical Learning Smooth Function Order Derivative Classical Learning

March 24, 2023

Double Descent Demystified: Identifying, Interpreting & Ablating the Sources of a Deep Learning Puzzle
Rylan Schaeffer, Mikail Khona, Zachary Robertson, Akhilan Boopathy, Kateryna Pistunova, Jason W. Rocks, Ila Rani Fiete, Oluwasanmi Koyejo
Deep Learning Machine Learning Linear Regression Source Table Double Descent Machine Translated Classical Learning Polynomial Regression

August 29, 2022

Approach of variable clustering and compression for learning large Bayesian networks
Anna V. Bubnova
Constructive Approach Linear Compression Cluster Analysis Classical Learning Block Local Learning

December 17, 2021

Provable Adversarial Robustness in the Quantum Model
Khashayar Barooti, Grzegorz Głuch, Ruediger Urbanke
Adversarial Robustness Quantum Learning Quantum Computation Quantum Model Classical Learning

Classical Learning

Papers

Fairness, Accuracy, and Unreliable Data

General Distribution Learning: A theoretical framework for Deep Learning

Separable Power of Classical and Quantum Learning Protocols Through the Lens of No-Free-Lunch Theorem

Generalization Error Analysis for Sparse Mixture-of-Experts: A Preliminary Study

Proper vs Improper Quantum PAC learning

On the Necessity of Metalearning: Learning Suitable Parameterizations for Learning Processes

Exact Count of Boundary Pieces of ReLU Classifiers: Towards the Proper Complexity Measure for Classification

How many samples are needed to leverage smoothness?

Double Descent Demystified: Identifying, Interpreting & Ablating the Sources of a Deep Learning Puzzle

Approach of variable clustering and compression for learning large Bayesian networks

Provable Adversarial Robustness in the Quantum Model