Constructive Reduction

Constructive reduction focuses on efficiently decreasing the size or complexity of data or models while preserving essential information. Current research explores this across diverse fields, employing techniques like deep learning frameworks for dimensionality reduction and continuous representation (e.g., in climate data analysis), information-theoretic approaches for pruning neural networks and causal model simplification, and optimal transport methods for data augmentation and dimensionality reduction. These advancements improve data storage, accelerate model inference, enhance explainability, and enable the application of complex models to resource-constrained environments, impacting fields ranging from climate science and anomaly detection to reinforcement learning and medical imaging.

Papers

October 24, 2024

Doubly Non-Central Beta Matrix Factorization for Stable Dimensionality Reduction of Bounded Support Matrix Data
Anjali N. Albert, Patrick Flaherty, Aaron Schein
Data Dimensionality Constructive Reduction Tucker Decomposition Matrix Decomposition DNA Methylation Support Estimation

October 18, 2024

Less is More: Selective Reduction of CT Data for Self-Supervised Pre-Training of Deep Learning Models with Contrastive Learning Improves Downstream Classification Performance
Daniel Wolf, Tristan Payer, Catharina Silvia Lisson, Christoph Gerhard Lisson, Meinrad Beer, Michael Götz, Timo Ropinski
Contrastive Learning Deep Learning Model Medical Image Constructive Reduction Contrastive Pre Training Downstream Classification

October 17, 2024

Align-ULCNet: Towards Low-Complexity and Robust Acoustic Echo and Noise Reduction
Shrishti Saha Shetu, Naveen Kumar Desiraju, Wolfgang Mack, Emanuël A. P. Habets
Industrial Disturbing Noise Constructive Reduction Low Complexity Noise Reduction Acoustic Echo U Net Model

October 9, 2024

Retrieval Replace Reduction: An effective visual token reduction method via semantic match
Yingen Liu, Fan Wu, Ruihui Li, Zhuo Tang, Kenli Li
Multimodal Large Language Model Multimodal Input Multimodal Task Constructive Reduction Visual Token Semantic Matching

September 25, 2024

Implicit Neural Representations for Simultaneous Reduction and Continuous Reconstruction of Multi-Altitude Climate Data
Alif Bin Abdul Qayyum, Xihaier Luo, Nathan M. Urban, Xiaoning Qian, Byung-Jun Yoon
Dimensionality Reduction Constructive Reduction Climate Data Conjunct ReSolution Wind Data

September 7, 2024

Reducing Events to Augment Log-based Anomaly Detection Models: An Empirical Study
Lingzhe Zhang, Tong Jia, Kangjin Wang, Mengxi Jia, Yang Yong, Ying Li
Anomaly Detection Empirical Study Event Description Constructive Reduction Anomaly Detection Model Log Based Anomaly Detection Log Pattern Noise Type Classification

August 23, 2024

Reduce, Reuse, Recycle: Categories for Compositional Reinforcement Learning
Georgios Bakirtzis, Michail Savvas, Ruihan Zhao, Sandeep Chinchali, Ufuk Topcu
Reinforcement Learning Markov Decision Process Constructive Reduction Category Shift Task Specification Reusable Data Exhaust Advanced Textile Recycling Compositional Reinforcement Learning

August 10, 2024

Structure and Reduction of MCTS for Explainable-AI
Ronit Bustin, Claudia V. Goldman
Monte Carlo Tree Search Inner Structure Sequential Decision Constructive Reduction Sequential Decision Making Problem Minimax McTs

July 24, 2024

Reduction of Nonlinear Distortion in Condenser Microphones Using a Simple Post-Processing Technique
Petr Honzík, Antonin Novak
Post Processing Constructive Reduction Reference Microphone Mem Microphone

June 4, 2024

Private Stochastic Convex Optimization with Heavy Tails: Near-Optimality from Simple Reductions
Hilal Asi, Daogao Liu, Kevin Tian
Near Optimality Constructive Reduction Private Algorithm Heavy Tailed Private Gradient Heavy Tail Approximate Differential Privacy Lipschitz Continuity Assumption

April 24, 2024

NEPENTHE: Entropy-Based Pruning as a Neural Network Depth's Reducer
Zhu Liao, Victor Quétu, Van-Tam Nguyen, Enzo Tartaglione
Neural Network Deep Neural Network Edge Pruning Constructive Reduction Depth Network Shannon Entropy

April 17, 2024

CORE: Data Augmentation for Link Prediction via Information Bottleneck
Kaiwen Dong, Zhichun Guo, Nitesh V. Chawla
Data Augmentation Graph Representation Learning Link Prediction Information Bottleneck Graph Data Constructive Reduction Novel Data Augmentation

April 2, 2024

A discussion about violin reduction: geometric analysis of contour lines and channel of minima
Philémon Beghin, Anne-Emmanuelle Ceulemans, François Glineur
Geometric Analysis Constructive Reduction Discussion Paper Musical Instrument Contour Error Morphological Exemplar

March 21, 2024

Reversible Jump Attack to Textual Classifiers with Modification Reduction
Mingze Ni, Zhensu Sun, Wei Liu
Adversarial Example Adversarial Sample Constructive Reduction

March 10, 2024

Risk-Sensitive RL with Optimized Certainty Equivalents via Reduction to Standard RL
Kaiwen Wang, Dawen Liang, Nathan Kallus, Wen Sun
Policy OpTimization Risk Sensitive Meta Reinforcement Learning Constructive Reduction Human Like RL Risk Sensitive Reinforcement Learning Certainty Equivalent

March 7, 2024

Strong Priority and Determinacy in Timed CCS
Luigi Liquori, Michael Mendler
Constructive Reduction Priority User Concurrency Model Transmission Determinacy

February 3, 2024

Self-Debiasing Large Language Models: Zero-Shot Recognition and Reduction of Stereotypes
Isabel O. Gallegos, Ryan A. Rossi, Joe Barrow, Md Mehrab Tanjim, Tong Yu, Hanieh Deilamsalehy, Ruiyi Zhang, Sungchul Kim, Franck Dernoncourt
Large Language Model Recognition Rate Bias Mitigation Stereotype Content Constructive Reduction Self Debiasing

January 13, 2024

Scalable and Efficient Methods for Uncertainty Estimation and Reduction in Deep Learning
Soyed Tuhin Ahmed
Deep Learning Variational Inference Uncertainty Estimation Bayesian Neural Network Potential Scalability Constructive Reduction Efficient Method Resistive Memory

November 30, 2023

Targeted Reduction of Causal Models
Armin Kekić, Bernhard Schölkopf, Michel Besserve
Causal Discovery Causal Representation Learning Constructive Reduction Causal Variable Interpretable Causal Complex Model

November 28, 2023

LightGaussian: Unbounded 3D Gaussian Compression with 15x Reduction and 200+ FPS
Zhiwen Fan, Kevin Wang, Kairun Wen, Zehao Zhu, Dejia Xu, Zhangyang Wang
Gaussian Splatting 3D Gaussian Neural Rendering 3d Representation Constructive Reduction

Constructive Reduction

Papers

Doubly Non-Central Beta Matrix Factorization for Stable Dimensionality Reduction of Bounded Support Matrix Data

Less is More: Selective Reduction of CT Data for Self-Supervised Pre-Training of Deep Learning Models with Contrastive Learning Improves Downstream Classification Performance

Align-ULCNet: Towards Low-Complexity and Robust Acoustic Echo and Noise Reduction

Retrieval Replace Reduction: An effective visual token reduction method via semantic match

Implicit Neural Representations for Simultaneous Reduction and Continuous Reconstruction of Multi-Altitude Climate Data

Reducing Events to Augment Log-based Anomaly Detection Models: An Empirical Study

Reduce, Reuse, Recycle: Categories for Compositional Reinforcement Learning

Structure and Reduction of MCTS for Explainable-AI

Reduction of Nonlinear Distortion in Condenser Microphones Using a Simple Post-Processing Technique

Private Stochastic Convex Optimization with Heavy Tails: Near-Optimality from Simple Reductions

NEPENTHE: Entropy-Based Pruning as a Neural Network Depth's Reducer

CORE: Data Augmentation for Link Prediction via Information Bottleneck

A discussion about violin reduction: geometric analysis of contour lines and channel of minima

Reversible Jump Attack to Textual Classifiers with Modification Reduction

Risk-Sensitive RL with Optimized Certainty Equivalents via Reduction to Standard RL

Strong Priority and Determinacy in Timed CCS

Self-Debiasing Large Language Models: Zero-Shot Recognition and Reduction of Stereotypes

Scalable and Efficient Methods for Uncertainty Estimation and Reduction in Deep Learning

Targeted Reduction of Causal Models

LightGaussian: Unbounded 3D Gaussian Compression with 15x Reduction and 200+ FPS