Deep Sets

In this paper, we study the problem of designing objective functions for machine learning problems defined on finite \emph{sets}. In contrast to traditional objective functions defined for machine learning problems operating on finite dimensional vectors, the new objective functions we propose are operating on finite sets and are invariant to permutations. Such problems are widespread, ranging from estimation of population statistics \citep{poczos13aistats}, via anomaly detection in piezometer data of embankment dams \citep{Jung15Exploration}, to cosmology \citep{Ntampaka16Dynamical,Ravanbakhsh16ICML1}. Our main theorem characterizes the permutation invariant objective functions and provides a family of functions to which any permutation invariant objective function must belong. This family of functions has a special structure which enables us to design a deep network architecture that can operate on sets and which can be deployed on a variety of scenarios including both unsupervised and supervised learning tasks. We demonstrate the applicability of our method on population statistic estimation, point cloud classification, set expansion, and image tagging.

• Publication year

  2017

• Venue

  Unknown venue

• Publication date

  2017-03-10

• Fields of study

  Mathematics, Computer Science

• Identifiers
  arXiv 1703.06114
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Deep Sets

  2017cited by this paper
• Learning a Probabilistic Latent Space of Object Shapes via 3D Generative-Adversarial Modeling

  2016cited by this paper
• Estimating Cosmological Parameters from the Dark Matter Distribution

  2016cited by this paper
• Generative and Discriminative Voxel Modeling with Convolutional Neural Networks

  2016cited by this paper
• Deep Learning

  2016cited by this paper
• Group Equivariant Convolutional Networks

  2016cited by this paper
• Fast and Accurate Deep Network Learning by Exponential Linear Units (ELUs)

  2015cited by this paper
• DeepPano: Deep Panoramic Representation for 3-D Shape Recognition

  2015cited by this paper
• Convolutional Networks on Graphs for Learning Molecular Fingerprints

  2015cited by this paper
• Multi-view Convolutional Neural Networks for 3D Shape Recognition

  2015cited by this paper
• ShapeNet: An Information-Rich 3D Model Repository

  2015cited by this paper
• DYNAMICAL MASS MEASUREMENTS OF CONTAMINATED GALAXY CLUSTERS USING MACHINE LEARNING

  2015cited by this paper
• Machine Learning Energies of 2 Million Elpasolite (ABC_{2}D_{6}) Crystals.

  2015cited by this paper
• Exploration and evaluation of AR, MPCA and KL anomaly detection techniques to embankment dam piezometer data

  2015cited by this paper
• VoxNet: A 3D Convolutional Neural Network for real-time object recognition

  2015cited by this paper
• 3D ShapeNets: A deep representation for volumetric shapes

  2014cited by this paper
• Deep Recursive Neural Networks for Compositionality in Language

  2014cited by this paper
• Learning Theory for Distribution Regression

  2014cited by this paper
• Microsoft COCO: Common Objects in Context

  2014cited by this paper
• Adam: A Method for Stochastic Optimization

  2014cited by this paper
• Deep Learning Face Attributes in the Wild

  2014cited by this paper
• Deep Symmetry Networks

  2014cited by this paper
• Recursive Deep Models for Semantic Compositionality Over a Sentiment Treebank

  2013cited by this paper
• Fast Image Tagging

  2013influential reference
• Domain Generalization via Invariant Feature Representation

  2013cited by this paper
• Distribution-Free Distribution Regression

  2013cited by this paper
• Distribution to Distribution Regression

  2013cited by this paper
• Distributed Representations of Words and Phrases and their Compositionality

  2013cited by this paper
• Spectral Networks and Locally Connected Networks on Graphs

  2013cited by this paper
• redMaPPer II: X-RAY AND SZ PERFORMANCE BENCHMARKS FOR THE SDSS CATALOG

  2013cited by this paper
• Learning from Distributions via Support Measure Machines

  2012cited by this paper
• Support Distribution Machines

  2012cited by this paper
• Tensor decompositions for learning latent variable models

  2012cited by this paper
• 3D is here: Point Cloud Library (PCL)

  2011cited by this paper
• TagProp: Discriminative metric learning in nearest neighbor models for image auto-annotation

  2009cited by this paper
• Latent Dirichlet Allocation

  2009cited by this paper
• A New Baseline for Image Annotation

  2008cited by this paper
• Analysis and evaluation of visual information systems performance

  2007cited by this paper
• Inference of population structure using multilocus genotype data: dominant markers and null alleles

  2007cited by this paper
• Bayesian Sets

  2005influential reference
• Training Invariant Support Vector Machines using Selective Sampling

  2005cited by this paper
• A mesh reconstruction algorithm driven by an intrinsic property of a point cloud

  2004cited by this paper
• Multiple Bernoulli relevance models for image and video annotation

  2004cited by this paper
• Labeling images with a computer game

  2004cited by this paper
• Max-Margin Markov Networks

  2003cited by this paper
• Inference of population structure using multilocus genotype data.

  2000cited by this paper
• Slicing Through Multicolor Space: Galaxy Redshifts from Broadband Photometry

  1995cited by this paper
• Interpolation of scattered data: Distance matrices and conditionally positive definite functions

  1986cited by this paper
• Galactic astronomy.

  1981cited by this paper

• Orthogonal Uplift Learning with Permutation-Invariant Representations for Combinatorial Treatments

  2026cites this paper
• The Ensemble Inverse Problem: Applications and Methods

  2026cites this paper
• MRFNet: Multi-reference fusion for image deblurring

  2026cites this paper
• MSMF-DBNet: Dual-Branch Model for Pulmonary Disease Recognition with Multi-Scale Feature Enhancement and Multichannel Fusion

  2026cites this paper
• Predicting the tabletability of binary mixtures from individual powder compaction behavior.

  2026cites this paper
• SMKC: Sketch Based Kernel Correlation Images for Variable Cardinality Time Series Anomaly Detection

  2026cites this paper
• Batch-of-Thought: Cross-Instance Learning for Enhanced LLM Reasoning

  2026cites this paper
• PanopTag: Simultaneously Tagging All Jets in a Particle Collision Event

  2026cites this paper
• Breaking Symmetry Bottlenecks in GNN Readouts

  2026cites this paper
• DeepTA: High-Speed Deep Camera Translation Averaging with Reverse Direction Invariance

  2026cites this paper
• A theoretical model of dynamical grammatical gender shifting based on set-valued set function

  2026cites this paper
• Hypernetwork-based approach for grid-independent functional data clustering

  2026cites this paper
• BayesFlow 2.0: Multi-Backend Amortized Bayesian Inference in Python

  2026cites this paper
• FAIRFORMER: A transformer architecture for discrete fair division

  2026cites this paper
• Robust soft sensing with causal and injectivity-preserving Graph Neural Network

  2026cites this paper
• Fermi Sets: Universal and interpretable neural architectures for fermions

  2026cites this paper
• Safe Heterogeneous Multi-Agent RL with Communication Regularization for Coordinated Target Acquisition

  2026cites this paper
• Principled Latent Diffusion for Graphs via Laplacian Autoencoders

  2026influential citation
• Collective Posterior Inference from Highly Variable Empirical Replicates

  2026cites this paper
• From independent patches to coordinated attention: Controlling information flow in vision transformers

  2026cites this paper
• Resilient Topology-Aware Coordination for Dynamic 3D UAV Networks under Node Failure

  2026cites this paper
• A Graph Meta-Network for Learning on Kolmogorov-Arnold Networks

  2026cites this paper
• Revisiting Integration of Image and Metadata for DICOM Series Classification: Cross-Attention and Dictionary Learning

  2026cites this paper
• Retrieving Patient-Specific Radiomic Feature Sets for Transparent Knee MRI Assessment

  2026cites this paper
• Distribution-Conditioned Transport

  2026cites this paper
• Accelerating Grokking through Spectral Analysis: A Multi-method Approach to Enhancing Abstract Rule Learning in Neural Networks

  2026cites this paper
• Fast Surrogate Learning for Multi-Objective UAV Placement in Motorway Intelligent Transportation System

  2026cites this paper
• Position: Message-passing and spectral GNNs are two sides of the same coin

  2026cites this paper
• Improving Set Function Approximation with Quasi-Arithmetic Neural Networks

  2026influential citation
• Learning Policy Representations for Steerable Behavior Synthesis

  2026influential citation
• High-Precision UAV Swarm Control: Evaluating Fractional-Order Methods for Close-Proximity Operations

  2026cites this paper
• Multi-Source Localization Based on Graph Representation Learning and Bayesian Optimization

  2026cites this paper
• Environment-Adaptive Covariate Selection: Learning When to Use Spurious Correlations for Out-of-Distribution Prediction

  2026influential citation
• Learning Resilient Elections with Adversarial GNNs

  2026cites this paper
• GenoPath-MCA: Multimodal masked cross-attention between genomics and pathology for survival prediction

  2026cites this paper
• A Statistical Assessment of Amortized Inference Under Signal-to-Noise Variation and Distribution Shift

  2026cites this paper
• A physics-informed graph neural network conserving linear and angular momentum for dynamical systems

  2026cites this paper
• Organ-Aware Attention Improves CT Triage and Classification

  2026cites this paper
• CPiRi: Channel Permutation-Invariant Relational Interaction for Multivariate Time Series Forecasting

  2026cites this paper
• Mean-Field Control on Sparse Graphs: From Local Limits to GNNs via Neighborhood Distributions

  2026cites this paper
• Self-supervised multi-stages set expansion for session-based recommendation

  2026influential citation
• Open Problems in Differentiable Social Choice: Learning Mechanisms, Decisions, and Alignment

  2026cites this paper
• Autoregressive Models for Knowledge Graph Generation

  2026cites this paper
• Probing Neural TSP Representations for Prescriptive Decision Support

  2026cites this paper
• MacNet: An End-to-End Manifold-Constrained Adaptive Clustering Network for Interpretable Whole Slide Image Classification

  2026cites this paper
• Efficient Point Cloud Processing with High-Dimensional Positional Encoding and Non-Local MLPs.

  2026cites this paper
• VersaTable: Generating Realistic Mixed-Type Tables Blending Categorical Tags, Variable-Length Lists, and Free Text

  2026cites this paper
• Unsupervised Continual Learning for Amortized Bayesian Inference

  2026cites this paper
• Symbol-Equivariant Recurrent Reasoning Models

  2026cites this paper
• Transform-Invariant Generative Ray Path Sampling for Efficient Radio Propagation Modeling

  2026cites this paper
• SELDON: Supernova Explosions Learned by Deep ODE Networks

  2026cites this paper
• Towards Improved Sentence Representations using Token Graphs

  2026influential citation
• The Density of Cross-Persistence Diagrams and Its Applications

  2026cites this paper
• HFF-BiG: Hierarchical feature fusion with bi-directional gating for 3D indoor instance segmentation

  2026cites this paper
• Quantitative Approximation Rates for Group Equivariant Learning

  2026influential citation
• PaReGTA: An LLM-based EHR Data Encoding Approach to Capture Temporal Information

  2026cites this paper
• The geometry of invariant learning: an information-theoretic analysis of data augmentation and generalization

  2026cites this paper
• Group-based learning on label-free phase-contrast images across dose and exposure time improves bioactive compound classification

  2026cites this paper
• CCTD-MARL: Coupled Communication-Task Decoupling Framework for Multi-Agent Systems Under Partial Observability

  2026cites this paper
• Semantically Labelled Automata for Multi-Task Reinforcement Learning with LTL Instructions

  2026cites this paper
• PAIR-Former: Budgeted Relational MIL for miRNA Target Prediction

  2026cites this paper
• Plain Transformers are Surprisingly Powerful Link Predictors

  2026cites this paper
• Quadrotor navigation considering attitude: A deep reinforcement learning method using tangent path rewards

  2026cites this paper
• Partially Permutation-Invariant Neural Network for Solving Two-Stage Stochastic AC-OPF Problem

  2026cites this paper
• db-ECBS: Interaction-Aware Multirobot Kinodynamic Motion Planning

  2026cites this paper
• Distributed penetration strategy for multi-USV based on scalable deep reinforcement learning

  2026cites this paper
• Consistency and Invariance Guided Multi-View Hypergraph Learning for Robust Hyperedge Prediction

  2026influential citation
• Toward Better Temporal Structures for Geopolitical Events Forecasting

  2026influential citation
• Learning from Historical Activations in Graph Neural Networks

  2026cites this paper
• From Mice to Trains: Amortized Bayesian Inference on Graph Data

  2026cites this paper
• Exact multi-parameter persistent homology of time-series data: Fast and variable topological inferences

  2026cites this paper
• DeepHalo: A Neural Choice Model with Controllable Context Effects

  2026cites this paper
• DeePM: Regime-Robust Deep Learning for Systematic Macro Portfolio Management

  2026cites this paper
• Revisiting Fine-Grained Image Analysis by Semantic-Part Alignment

  2026cites this paper
• On the Kolmogorov Superposition Theorem and Regular Means

  2026cites this paper
• Multi-scale local geometry feature and global context learning with Kolmogorov-Arnold representation for 3D semantic segmentation

  2026cites this paper
• Reasoning While Recommending: Entropy-Guided Latent Reasoning in Generative Re-ranking Models

  2026cites this paper
• Two-Layer Reinforcement Learning-Assisted Joint Beamforming and Trajectory Optimization for Multi-UAV Downlink Communications

  2026cites this paper
• Permutation-Invariant Physics-Informed Neural Network for Region-to-Region Sound Field Reconstruction

  2026cites this paper
• Neural Cooperative Reach-While-Avoid Certificates for Interconnected Systems

  2026cites this paper
• A Probabilistic Neuro-symbolic Layer for Algebraic Constraint Satisfaction

  2025cites this paper
• Multi-task Learning of Classification and Generation for Set-structured Data

  2025cites this paper
• Multimodal multi-instance evidence fusion neural networks for cancer survival prediction

  2025cites this paper
• Real-time probabilistic model updating and damage detection using Machine Learning-based likelihood-free inference

  2025cites this paper
• Reinforcement Learning-based Heuristics to Guide Domain-Independent Dynamic Programming

  2025cites this paper
• TeMP-TraG: Edge-based Temporal Message Passing in Transaction Graphs

  2025cites this paper
• Symmetry-Informed Graph Neural Networks for Carbon Dioxide Isotherm and Adsorption Prediction in Aluminum-Substituted Zeolites

  2025cites this paper
• SPECTra: Scalable Multi-Agent Reinforcement Learning with Permutation-Free Networks

  2025cites this paper
• A Graph-Based Framework for Interpretable Whole Slide Image Analysis

  2025cites this paper
• Enhancing job salary prediction with disentangled composition effect modeling: a neural prototyping approach

  2025cites this paper
• Edge-Centric Embeddings of Digraphs: Properties and Stability Under Sparsification

  2025cites this paper
• Position-aware Hypergraph Message-Passing Neural Network

  2025cites this paper
• Permutation Equivariant Neural Networks for Symmetric Tensors

  2025influential citation
• Tokenize Image as a Set

  2025cites this paper
• A QUBO Framework for Team Formation

  2025cites this paper
• PointNSP: Autoregressive 3D Point Cloud Generation with Next-Scale Level-of-Detail Prediction

  2025cites this paper
• A Robust Attention-based Cardinality Estimator with Domain Adaptation

  2025cites this paper
• Overlap-aware meta-learning attention to enhance hypergraph neural networks for node classification

  2025cites this paper
• ADR-SALD: Attention-Based Deep Residual Sign Agnostic Learning With Derivatives for Implicit Surface Reconstruction

  2025cites this paper
• Transferable foundation models for geometric tasks on point cloud representations: geometric neural operators

  2025cites this paper