Search Results for author: Santiago Segarra

Found 61 papers, 30 papers with code

Windowed Fourier Analysis for Signal Processing on Graph Bundles

no code implementations11 Feb 2023 T. Mitchell Roddenberry, Santiago Segarra

We consider the task of representing signals supported on graph bundles, which are generalizations of product graphs that allow for "twists" in the product structure.


Unsupervised Learning of Sampling Distributions for Particle Filters

no code implementations2 Feb 2023 Fernando Gama, Nicolas Zilberstein, Martin Sevilla, Richard Baraniuk, Santiago Segarra

Thus, the crux of particle filters lies in designing sampling distributions that are both easy to sample from and lead to accurate estimators.

Design Synthesis

Joint graph learning from Gaussian observations in the presence of hidden nodes

1 code implementation4 Dec 2022 Samuel Rey, Madeline Navarro, Andrei Buciulea, Santiago Segarra, Antonio G. Marques

Motivated by this, we propose a joint graph learning method that takes into account the presence of hidden (latent) variables.

Graph Learning Graph Similarity

Delay-aware Backpressure Routing Using Graph Neural Networks

no code implementations19 Nov 2022 Zhongyuan Zhao, Bojan Radojicic, Gunjan Verma, Ananthram Swami, Santiago Segarra

In this work, we improve upon the widely-used metric of hop distance (and its variants) for the shortest path bias by introducing a bias based on the link duty cycle, which we predict using a graph convolutional neural network.

Graph Filters for Signal Processing and Machine Learning on Graphs

no code implementations16 Nov 2022 Elvin Isufi, Fernando Gama, David I. Shuman, Santiago Segarra

For time series and image data that reside on Euclidean domains, filters are the crux of many signal processing and machine learning techniques, including convolutional neural networks.

Time Series Analysis

Beyond Hawkes: Neural Multi-event Forecasting on Spatio-temporal Point Processes

1 code implementation5 Nov 2022 Negar Erfanian, Santiago Segarra, Maarten de Hoop

Predicting discrete events in time and space has many scientific applications, such as predicting hazardous earthquakes and outbreaks of infectious diseases.

Point Processes

GraphMAD: Graph Mixup for Data Augmentation using Data-Driven Convex Clustering

1 code implementation27 Oct 2022 Madeline Navarro, Santiago Segarra

Mixup is a data augmentation method to create new training data by linearly interpolating between pairs of data samples and their labels.

Data Augmentation Graph Classification

Accelerated massive MIMO detector based on annealed underdamped Langevin dynamics

1 code implementation26 Oct 2022 Nicolas Zilberstein, Chris Dick, Rahman Doost-Mohammady, Ashutosh Sabharwal, Santiago Segarra

We propose a multiple-input multiple-output (MIMO) detector based on an annealed version of the \emph{underdamped} Langevin (stochastic) dynamic.

Joint Network Topology Inference via a Shared Graphon Model

1 code implementation17 Sep 2022 Madeline Navarro, Santiago Segarra

The proposed joint network and graphon estimation is further enhanced with the introduction of a robust method for noisy graph sampling information.

Graphon Estimation Graph Sampling

Hypergraphs with Edge-Dependent Vertex Weights: p-Laplacians and Spectral Clustering

1 code implementation15 Aug 2022 Yu Zhu, Santiago Segarra

We study p-Laplacians and spectral clustering for a recently proposed hypergraph model that incorporates edge-dependent vertex weights (EDVW).

Enhanced graph-learning schemes driven by similar distributions of motifs

no code implementations11 Jul 2022 Samuel Rey, T. Mitchell Roddenberry, Santiago Segarra, Antonio G. Marques

Guided by this, we first assume that we have a reference graph that is related to the sought graph (in the sense of having similar motif densities) and then, we exploit this relation by incorporating a similarity constraint and a regularization term in the network topology inference optimization problem.

Graph Learning Inference Optimization

Annealed Langevin Dynamics for Massive MIMO Detection

1 code implementation11 May 2022 Nicolas Zilberstein, Chris Dick, Rahman Doost-Mohammady, Ashutosh Sabharwal, Santiago Segarra

Based on the proposed MIMO detector, we also design a robust version of the method by unfolding and parameterizing one term -- the score of the likelihood -- by a neural network.

Distributed Link Sparsification for Scalable Scheduling Using Graph Neural Networks

1 code implementation27 Mar 2022 Zhongyuan Zhao, Ananthram Swami, Santiago Segarra

Distributed scheduling algorithms for throughput or utility maximization in dense wireless multi-hop networks can have overwhelmingly high overhead, causing increased congestion, energy consumption, radio footprint, and security vulnerability.


Detection by Sampling: Massive MIMO Detector based on Langevin Dynamics

1 code implementation24 Feb 2022 Nicolas Zilberstein, Chris Dick, Rahman Doost-Mohammady, Ashutosh Sabharwal, Santiago Segarra

Optimal symbol detection in multiple-input multiple-output (MIMO) systems is known to be an NP-hard problem.

On Local Distributions in Graph Signal Processing

no code implementations22 Feb 2022 T. Mitchell Roddenberry, Fernando Gama, Richard G. Baraniuk, Santiago Segarra

Leveraging this, we are able to seamlessly compare graphs of different sizes and coming from different models, yielding results on the convergence of spectral densities, transferability of filters across arbitrary graphs, and continuity of graph signal properties with respect to the distribution of local substructures.

Graphon-aided Joint Estimation of Multiple Graphs

1 code implementation11 Feb 2022 Madeline Navarro, Santiago Segarra

We consider the problem of estimating the topology of multiple networks from nodal observations, where these networks are assumed to be drawn from the same (unknown) random graph model.

Graphon Estimation

Graph-based Algorithm Unfolding for Energy-aware Power Allocation in Wireless Networks

no code implementations27 Jan 2022 Boning Li, Gunjan Verma, Santiago Segarra

We develop a novel graph-based trainable framework to maximize the weighted sum energy efficiency (WSEE) for power allocation in wireless communication networks.

Power Allocation for Wireless Federated Learning using Graph Neural Networks

2 code implementations15 Nov 2021 Boning Li, Ananthram Swami, Santiago Segarra

We propose a data-driven approach for power allocation in the context of federated learning (FL) over interference-limited wireless networks.

Federated Learning

Stability Analysis of Unfolded WMMSE for Power Allocation

1 code implementation14 Oct 2021 Arindam Chowdhury, Fernando Gama, Santiago Segarra

Power allocation is one of the fundamental problems in wireless networks and a wide variety of algorithms address this problem from different perspectives.

Robust MIMO Detection using Hypernetworks with Learned Regularizers

1 code implementation13 Oct 2021 Nicolas Zilberstein, Chris Dick, Rahman Doost-Mohammady, Ashutosh Sabharwal, Santiago Segarra

Our method is based on hypernetworks that generate the parameters of a neural network-based detector that works well on a specific channel.

Label Propagation across Graphs: Node Classification using Graph Neural Tangent Kernels

no code implementations7 Oct 2021 Artun Bayer, Arindam Chowdhury, Santiago Segarra

In this context, our current work considers a challenging inductive setting where a set of labeled graphs are available for training while the unlabeled target graph is completely separate, i. e., there are no connections between labeled and unlabeled nodes.

Node Classification

Joint inference of multiple graphs with hidden variables from stationary graph signals

1 code implementation5 Oct 2021 Samuel Rey, Andrei Buciulea, Madeline Navarro, Santiago Segarra, Antonio G. Marques

Learning graphs from sets of nodal observations represents a prominent problem formally known as graph topology inference.

A Robust Alternative for Graph Convolutional Neural Networks via Graph Neighborhood Filters

1 code implementation2 Oct 2021 Victor M. Tenorio, Samuel Rey, Fernando Gama, Santiago Segarra, Antonio G. Marques

Graph convolutional neural networks (GCNNs) are popular deep learning architectures that, upon replacing regular convolutions with graph filters (GFs), generalize CNNs to irregular domains.

Denoising Node Classification

Untrained Graph Neural Networks for Denoising

1 code implementation24 Sep 2021 Samuel Rey, Santiago Segarra, Reinhard Heckel, Antonio G. Marques

This paper introduces two untrained graph neural network architectures for graph signal denoising, provides theoretical guarantees for their denoising capabilities in a simple setup, and numerically validates the theoretical results in more general scenarios.


Hodgelets: Localized Spectral Representations of Flows on Simplicial Complexes

no code implementations17 Sep 2021 T. Mitchell Roddenberry, Florian Frantzen, Michael T. Schaub, Santiago Segarra

We first show that the Hodge Laplacian can be used in lieu of the graph Laplacian to construct a family of wavelets for higher-order signals on simplicial complexes.

Link Scheduling using Graph Neural Networks

1 code implementation12 Sep 2021 Zhongyuan Zhao, Gunjan Verma, Chirag Rao, Ananthram Swami, Santiago Segarra

Test results on medium-sized wireless networks show that our centralized heuristic can reach a near-optimal solution quickly, and our distributed heuristic based on a shallow GCN can reduce by nearly half the suboptimality gap of the distributed greedy solver with minimal increase in complexity.


Robust Hierarchical Clustering for Directed Networks: An Axiomatic Approach

no code implementations16 Aug 2021 Gunnar Carlsson, Facundo Mémoli, Santiago Segarra

We begin by introducing three practical properties associated with the notion of robustness in hierarchical clustering: linear scale preservation, stability, and excisiveness.

Signal processing on simplicial complexes

no code implementations14 Jun 2021 Michael T. Schaub, Jean-Baptiste Seby, Florian Frantzen, T. Mitchell Roddenberry, Yu Zhu, Santiago Segarra

Higher-order networks have so far been considered primarily in the context of studying the structure of complex systems, i. e., the higher-order or multi-way relations connecting the constituent entities.

Denoising Time Series Analysis

Graph-signal Reconstruction and Blind Deconvolution for Structured Inputs

no code implementations31 May 2021 David Ramírez, Antonio G. Marques, Santiago Segarra

When either the input or the filter coefficients are known, this is tantamount to assuming that the signals of interest live on a subspace defined by the supporting graph.

An Impossibility Theorem for Node Embedding

no code implementations27 May 2021 T. Mitchell Roddenberry, Yu Zhu, Santiago Segarra

With the increasing popularity of graph-based methods for dimensionality reduction and representation learning, node embedding functions have become important objects of study in the literature.

Dimensionality Reduction Representation Learning

Free Energy Node Embedding via Generalized Skip-gram with Negative Sampling

1 code implementation19 May 2021 Yu Zhu, Ananthram Swami, Santiago Segarra

On the other hand, we propose a matrix factorization method based on a loss function that generalizes that of the skip-gram model with negative sampling to arbitrary similarity matrices.

Link Prediction Node Classification +1

GIST: Distributed Training for Large-Scale Graph Convolutional Networks

1 code implementation20 Feb 2021 Cameron R. Wolfe, Jingkang Yang, Arindam Chowdhury, Chen Dun, Artun Bayer, Santiago Segarra, Anastasios Kyrillidis

The graph convolutional network (GCN) is a go-to solution for machine learning on graphs, but its training is notoriously difficult to scale both in terms of graph size and the number of model parameters.

BIG-bench Machine Learning Graph Sampling

Co-clustering Vertices and Hyperedges via Spectral Hypergraph Partitioning

1 code implementation19 Feb 2021 Yu Zhu, Boning Li, Santiago Segarra

We propose a novel method to co-cluster the vertices and hyperedges of hypergraphs with edge-dependent vertex weights (EDVWs).

hypergraph partitioning

Principled Simplicial Neural Networks for Trajectory Prediction

2 code implementations19 Feb 2021 T. Mitchell Roddenberry, Nicholas Glaze, Santiago Segarra

We consider the construction of neural network architectures for data on simplicial complexes.

Trajectory Prediction

Signal Processing on Higher-Order Networks: Livin' on the Edge ... and Beyond

no code implementations14 Jan 2021 Michael T. Schaub, Yu Zhu, Jean-Baptiste Seby, T. Mitchell Roddenberry, Santiago Segarra

In the context of simplicial complexes, we specifically focus on signal processing using the Hodge Laplacian matrix, a multi-relational operator that leverages the special structure of simplicial complexes and generalizes desirable properties of the Laplacian matrix in graph signal processing.


Blind Demixing of Diffused Graph Signals

no code implementations24 Dec 2020 Fernando J. Iglesias Garcia, Santiago Segarra, Antonio G. Marques

Using graphs to model irregular information domains is an effective approach to deal with some of the intricacies of contemporary (network) data.

Rank-One Measurements of Low-Rank PSD Matrices Have Small Feasible Sets

no code implementations17 Dec 2020 T. Mitchell Roddenberry, Santiago Segarra, Anastasios Kyrillidis

We study the role of the constraint set in determining the solution to low-rank, positive semidefinite (PSD) matrix sensing problems.

Adaptive Contention Window Design using Deep Q-learning

1 code implementation18 Nov 2020 Abhishek Kumar, Gunjan Verma, Chirag Rao, Ananthram Swami, Santiago Segarra

We study the problem of adaptive contention window (CW) design for random-access wireless networks.


Distributed Scheduling using Graph Neural Networks

1 code implementation18 Nov 2020 Zhongyuan Zhao, Gunjan Verma, Chirag Rao, Ananthram Swami, Santiago Segarra

In small- to middle-sized wireless networks with tens of links, even a shallow GCN-based MWIS scheduler can leverage the topological information of the graph to reduce in half the suboptimality gap of the distributed greedy solver with good generalizability across graphs and minimal increase in complexity.


Network topology change-point detection from graph signals with prior spectral signatures

no code implementations21 Oct 2020 Chiraag Kaushik, T. Mitchell Roddenberry, Santiago Segarra

We assume that signals on the nodes of the graph are regularized by the underlying graph structure via a graph filtering model, which we then leverage to distill the graph topology change-point detection problem to a subspace detection problem.

Change Point Detection

Joint Inference of Multiple Graphs from Matrix Polynomials

no code implementations16 Oct 2020 Madeline Navarro, Yuhao Wang, Antonio G. Marques, Caroline Uhler, Santiago Segarra

Inferring graph structure from observations on the nodes is an important and popular network science task.

Network Topology Inference with Graphon Spectral Penalties

no code implementations15 Oct 2020 T. Mitchell Roddenberry, Madeline Navarro, Santiago Segarra

In particular, we consider the case where the graph was drawn from a graphon model, and we supplement our convex optimization problem with a provably-valid regularizer on the spectrum of the graph to be recovered.

Unfolding WMMSE using Graph Neural Networks for Efficient Power Allocation

1 code implementation22 Sep 2020 Arindam Chowdhury, Gunjan Verma, Chirag Rao, Ananthram Swami, Santiago Segarra

We study the problem of optimal power allocation in a single-hop ad hoc wireless network.

Signal Processing on Directed Graphs

no code implementations2 Aug 2020 Antonio G. Marques, Santiago Segarra, Gonzalo Mateos

This paper provides an overview of the current landscape of signal processing (SP) on directed graphs (digraphs).

Causal Inference

An Underparametrized Deep Decoder Architecture for Graph Signals

1 code implementation2 Aug 2019 Samuel Rey, Antonio G. Marques, Santiago Segarra

While deep convolutional architectures have achieved remarkable results in a gamut of supervised applications dealing with images and speech, recent works show that deep untrained non-convolutional architectures can also outperform state-of-the-art methods in several tasks such as image compression and denoising.

Denoising Image Compression

Blind identification of stochastic block models from dynamical observations

no code implementations22 May 2019 Michael T. Schaub, Santiago Segarra, John N. Tsitsiklis

We consider a blind identification problem in which we aim to recover a statistical model of a network without knowledge of the network's edges, but based solely on nodal observations of a certain process.

Stochastic Block Model

Graph-based Semi-Supervised & Active Learning for Edge Flows

1 code implementation17 May 2019 Junteng Jia, Michael T. Schaub, Santiago Segarra, Austin R. Benson

The first strategy selects edges to minimize the reconstruction error bound and works well on flows that are approximately divergence-free.

Active Learning

Spectral partitioning of time-varying networks with unobserved edges

no code implementations26 Apr 2019 Michael T. Schaub, Santiago Segarra, Hoi-To Wai

We discuss a variant of `blind' community detection, in which we aim to partition an unobserved network from the observation of a (dynamical) graph signal defined on the network.

Community Detection

Blind Community Detection from Low-rank Excitations of a Graph Filter

no code implementations5 Sep 2018 Hoi-To Wai, Santiago Segarra, Asuman E. Ozdaglar, Anna Scaglione, Ali Jadbabaie

The paper shows that communities can be detected by applying a spectral method to the covariance matrix of graph signals.

Community Detection

Centrality measures for graphons: Accounting for uncertainty in networks

no code implementations28 Jul 2017 Marco Avella-Medina, Francesca Parise, Michael T. Schaub, Santiago Segarra

Using the theory of linear integral operators, we define degree, eigenvector, Katz and PageRank centrality functions for graphons and establish concentration inequalities demonstrating that graphon centrality functions arise naturally as limits of their counterparts defined on sequences of graphs of increasing size.

Stylometric Analysis of Early Modern Period English Plays

no code implementations18 Oct 2016 Mark Eisen, Santiago Segarra, Gabriel Egan, Alejandro Ribeiro

We first study the similarity of writing styles between Early English playwrights by comparing the profile WANs.

Hierarchical Clustering of Asymmetric Networks

no code implementations21 Jul 2016 Gunnar Carlsson, Facundo Mémoli, Alejandro Ribeiro, Santiago Segarra

This paper considers networks where relationships between nodes are represented by directed dissimilarities.

Admissible Hierarchical Clustering Methods and Algorithms for Asymmetric Networks

no code implementations21 Jul 2016 Gunnar Carlsson, Facundo Mémoli, Alejandro Ribeiro, Santiago Segarra

This paper characterizes hierarchical clustering methods that abide by two previously introduced axioms -- thus, denominated admissible methods -- and proposes tractable algorithms for their implementation.

Excisive Hierarchical Clustering Methods for Network Data

no code implementations21 Jul 2016 Gunnar Carlsson, Facundo Mémoli, Alejandro Ribeiro, Santiago Segarra

We introduce two practical properties of hierarchical clustering methods for (possibly asymmetric) network data: excisiveness and linear scale preservation.

Authorship Attribution through Function Word Adjacency Networks

no code implementations17 Jun 2014 Santiago Segarra, Mark Eisen, Alejandro Ribeiro

Attribution accuracy is observed to exceed the one achieved by methods that rely on word frequencies alone.

Hierarchical Quasi-Clustering Methods for Asymmetric Networks

no code implementations17 Apr 2014 Gunnar Carlsson, Facundo Mémoli, Alejandro Ribeiro, Santiago Segarra

This paper introduces hierarchical quasi-clustering methods, a generalization of hierarchical clustering for asymmetric networks where the output structure preserves the asymmetry of the input data.

Axiomatic Construction of Hierarchical Clustering in Asymmetric Networks

no code implementations31 Jan 2013 Gunnar Carlsson, Facundo Mémoli, Alejandro Ribeiro, Santiago Segarra

Our construction of hierarchical clustering methods is based on defining admissible methods to be those methods that abide by the axioms of value - nodes in a network with two nodes are clustered together at the maximum of the two dissimilarities between them - and transformation - when dissimilarities are reduced, the network may become more clustered but not less.

Cannot find the paper you are looking for? You can Submit a new open access paper.