Marcus Poggi
Marcus Poggi has a total of 33 co-authors. Here are the top co-authors by number of shared publications:
- Eduardo Uchoa
Departamento de Engenharia de Produção Universidade Federal Fluminense Rua Passo da Patria 156 24210-240 Niteroi Rio de Janeiro Brazil
7 shared publications - Artur Alves Pessoa
Departamento de Engenharia de Produção Universidade Federal Fluminense Rua Passo da Patria 156 24210-240 Niteroi Rio de Janeiro Brazil
4 shared publications - Thibaut Vidal
CIRRELT & SCALE-AI Chair in Data-Driven Supply Chains, Department of Mathematics and Industrial Engineering, École Polytechnique de Montréal, Montréal, Quebec H3T 1J4, Canada
4 shared publications - Jens Lysgaard
Aarhus School of Business
2 shared publications - Ricardo Fukasawa
GeorgiaTech
2 shared publications - Alexandre M. Florio
Last Mile Planning, Amazon, Bellevue, Washington 98004
1 shared publication
Latest Publications
- 8 May 2024
- Published by Springer Nature in SN Computer Science
- Vol. 5 (5) , 1-27
Creating Automatic Connections for Personal Knowledge Management
- 8 May 2024
- Published by Springer Nature in SN Computer Science
- Vol. 5 (5) , 1-27
Creating Automatic Connections for Personal Knowledge Management
- Open Access
Regularization and Optimization in Model-Based Clustering
Due to their conceptual simplicity, k-means algorithm variants have been extensively used for unsupervised cluster analysis. However, one main shortcoming of these algorithms is that they essentially fit a mixture of identical spherical Gaussians to data that vastly deviates from such a distribution. In comparison, general Gaussian Mixture Models (GMMs) can fit richer structures but require estimating a quadratic number of parameters per cluster to represent the covariance matrices. This poses two main issues: (i) the underlying optimization problems are challenging due to their larger number of local minima, and (ii) their solutions can overfit the data. In this work, we design search strategies that circumvent both issues. We develop more effective optimization algorithms for general GMMs, and we combine these algorithms with regularization strategies that avoid overfitting. Through extensive computational analyses, we observe that optimization or regularization in isolation does not substantially improve cluster recovery. However, combining these techniques permits a completely new level of performance previously unachieved by k-means algorithm variants, unraveling vastly different cluster structures. These results shed new light on the current status quo between GMM and k-means methods and suggest the more frequent use of general GMMs for data exploration. To facilitate such applications, we provide open-source code as well as Julia packages (UnsupervisedClustering.jl and RegularizedCovarianceMatrices.jl) implementing the proposed techniques.
Open AccessRegularization and Optimization in Model-Based Clustering
Due to their conceptual simplicity, k-means algorithm variants have been extensively used for unsupervised cluster analysis. However, one main shortcoming of these algorithms is that they essentially fit a mixture of identical spherical Gaussians to data that vastly deviates from such a distribution. In comparison, general Gaussian Mixture Models (GMMs) can fit richer structures but require estimating a quadratic number of parameters per cluster to represent the covariance matrices. This poses two main issues: (i) the underlying optimization problems are challenging due to their larger number of local minima, and (ii) their solutions can overfit the data. In this work, we design search strategies that circumvent both issues. We develop more effective optimization algorithms for general GMMs, and we combine these algorithms with regularization strategies that avoid overfitting. Through extensive computational analyses, we observe that optimization or regularization in isolation does not substantially improve cluster recovery. However, combining these techniques permits a completely new level of performance previously unachieved by k-means algorithm variants, unraveling vastly different cluster structures. These results shed new light on the current status quo between GMM and k-means methods and suggest the more frequent use of general GMMs for data exploration. To facilitate such applications, we provide open-source code as well as Julia packages (UnsupervisedClustering.jl and RegularizedCovarianceMatrices.jl) implementing the proposed techniques.
- 14 September 2023
- Published by Springer Nature
- p. 388-401
Binary Matrix Factorization Discretization
- 1 September 2022
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Transportation Science
- Vol. 56 (5) , 1393-1408
Vehicle Routing with Stochastic Demands and Partial Reoptimization
We consider the vehicle routing problem with stochastic demands (VRPSD), a problem in which customer demands are known in distribution at the route planning stage and revealed during route execution upon arrival at each customer. A long-standing open question in the VRPSD concerns the benefits of allowing, during route execution, partial reordering of the planned customer visits. Given the practical importance of this question and the growing interest on the VRPSD under optimal restocking, we study the VRPSD under a recourse policy known as the switch policy. The switch policy is a canonical reoptimization policy that permits only pairs of successive customers to be reordered. We consider this policy jointly with optimal preventive restocking and introduce a branch-cut-and-price algorithm to compute optimal a priori routing plans in this context. At its core, this algorithm features pricing routines where value functions represent the expected cost-to-go along planned routes for all possible states and reordering decisions. To ensure pricing tractability, we adopt a strategy that combines elementary pricing with completion bounds of varying complexity, and we solve the pricing problem without relying on dominance rules. Our numerical experiments demonstrate the effectiveness of the algorithm for solving instances with up to 50 customers. Notably, they also give us new insights into the value of reoptimization. The switch policy enables significant cost savings over optimal restocking when the planned routes come from an algorithm built on a deterministic approximation of the data, an important scenario given the difficulty of finding optimal VRPSD solutions. The benefits are smaller when comparing optimal a priori VRPSD solutions obtained for both recourse policies. As it appears, further cost savings may require joint reordering and reassignment of customer visits among vehicles when the context permits.
- 1 September 2022
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Transportation Science
- Vol. 56 (5) , 1393-1408
Vehicle Routing with Stochastic Demands and Partial Reoptimization
We consider the vehicle routing problem with stochastic demands (VRPSD), a problem in which customer demands are known in distribution at the route planning stage and revealed during route execution upon arrival at each customer. A long-standing open question in the VRPSD concerns the benefits of allowing, during route execution, partial reordering of the planned customer visits. Given the practical importance of this question and the growing interest on the VRPSD under optimal restocking, we study the VRPSD under a recourse policy known as the switch policy. The switch policy is a canonical reoptimization policy that permits only pairs of successive customers to be reordered. We consider this policy jointly with optimal preventive restocking and introduce a branch-cut-and-price algorithm to compute optimal a priori routing plans in this context. At its core, this algorithm features pricing routines where value functions represent the expected cost-to-go along planned routes for all possible states and reordering decisions. To ensure pricing tractability, we adopt a strategy that combines elementary pricing with completion bounds of varying complexity, and we solve the pricing problem without relying on dominance rules. Our numerical experiments demonstrate the effectiveness of the algorithm for solving instances with up to 50 customers. Notably, they also give us new insights into the value of reoptimization. The switch policy enables significant cost savings over optimal restocking when the planned routes come from an algorithm built on a deterministic approximation of the data, an important scenario given the difficulty of finding optimal VRPSD solutions. The benefits are smaller when comparing optimal a priori VRPSD solutions obtained for both recourse policies. As it appears, further cost savings may require joint reordering and reassignment of customer visits among vehicles when the context permits.
- 7 December 2020
- Published by Sociedade Brasileira de Computação
- p. 672-685
DASRS Rest: Um Algoritmo Eficiente de Detecção de Anomalias em Tempo Real para Data Centers
A grande quantidade de equipamentos, com diferentes configurações, e atualizações constantes de software e hardware em data centers, tornam difícil o uso de sistemas de monitoração baseados na configuração de thresholds. Este artigo propõe utilizar detecção de anomalias para prevenção de falhas e apresenta o algoritmo Decreased Anomaly Score by Repeated Sequence (DASRS) Rest, que detecta anomalias sem exigir conhecimento prévio do serviço monitorado. Avaliamos o desempenho do DASRS Rest utilizando o framework Numenta Anomaly Benchmark (NAB). O algoritmo proposto possui bons resultados de acurácia, o menor consumo de memória e é o mais rápido, quando comparado com diversos algoritmos do estado da arte.
- 7 December 2020
- Published by Sociedade Brasileira de Computação
- p. 672-685
DASRS Rest: Um Algoritmo Eficiente de Detecção de Anomalias em Tempo Real para Data Centers
A grande quantidade de equipamentos, com diferentes configurações, e atualizações constantes de software e hardware em data centers, tornam difícil o uso de sistemas de monitoração baseados na configuração de thresholds. Este artigo propõe utilizar detecção de anomalias para prevenção de falhas e apresenta o algoritmo Decreased Anomaly Score by Repeated Sequence (DASRS) Rest, que detecta anomalias sem exigir conhecimento prévio do serviço monitorado. Avaliamos o desempenho do DASRS Rest utilizando o framework Numenta Anomaly Benchmark (NAB). O algoritmo proposto possui bons resultados de acurácia, o menor consumo de memória e é o mais rápido, quando comparado com diversos algoritmos do estado da arte.
- 1 July 2019
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in INFORMS Journal on Computing
- Vol. 31 (3) , 527-543
Two-Dimensional Phase Unwrapping via Balanced Spanning Forests
Phase unwrapping is the process of recovering a continuous phase signal from an original signal wrapped in the (−π,π] interval. It is a critical step of coherent signal processing, with application...
- 1 July 2019
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in INFORMS Journal on Computing
- Vol. 31 (3) , 527-543
Two-Dimensional Phase Unwrapping via Balanced Spanning Forests
Phase unwrapping is the process of recovering a continuous phase signal from an original signal wrapped in the (−π,π] interval. It is a critical step of coherent signal processing, with application...
- 27 October 2017
- Published by Springer Nature in Journal of Global Optimization
- Vol. 72 (1) , 27-53
The minimum distance superset problem: formulations and algorithms
- 27 October 2017
- Published by Springer Nature in Journal of Global Optimization
- Vol. 72 (1) , 27-53
The minimum distance superset problem: formulations and algorithms
- 16 February 2017
- Published by Springer Nature in Optimization and Engineering
- Vol. 18 (1) , 215-240
Petroleum supply planning: reformulations and a novel decomposition algorithm
- 16 February 2017
- Published by Springer Nature in Optimization and Engineering
- Vol. 18 (1) , 215-240
Petroleum supply planning: reformulations and a novel decomposition algorithm
- 1 January 2017
- Published by Elsevier in European Journal of Operational Research
- Vol. 256 (1) , 252-260
A more human-like portfolio optimization approach
- 1 January 2017
- Published by Elsevier in European Journal of Operational Research
- Vol. 256 (1) , 252-260
A more human-like portfolio optimization approach
- Open Access
- 1 December 2015
- Published by Springer Nature
- p. 45-58
Adaptive Automated Storytelling Based on Audience Response
- 1 December 2015
- Published by Springer Nature
- p. 45-58
Open AccessAdaptive Automated Storytelling Based on Audience Response