Mihaela Elena Breaban

Home
Research interests
Publications
For students [RO]
  • R Miron, R. Albert, M Breaban. A Two-Stage Atrous Convolution Neural Network for Brain Tumor Segmentation and Survival Prediction, in Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries, BrainLesion workshop within MICCAI 2020, published 2021.
  • R Miron, C Moisii, M Breaban. Revealing Lung Affections from CTs. A Comparative Analysis of Various Deep Learning Approaches for Dealing with Volumetric Datat, Working Notes of CLEF 2020 - Conference and Labs of the Evaluation Forum, 2020.
  • RM Ungureanu ME Breaban. An Evolutionary Distributed Framework for Projection Pursuit, Procedia Computer Science 176, 2020.
  • VI Lupoaie, IA Chili, ME Breaban, M Raschip. SOM-Guided Evolutionary Search for Solving MinMax Multiple-TSP, IEEE Congress on Evolutionary Computation (CEC), 73-80, 2019.
  • C Padurariu, ME Breaban. Dealing with Data Imbalance in Text Classification, Procedia Computer Science 159, 736-745, 2019.
  • D Minzat, M Breaban, H Luchian. Modeling real estate dynamics using survival analysis, IEEE 20th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing (SYNASC), 215-222, 2018.
  • R Necula, M Raschip, M Breaban. Balancing the Subtours for Multiple TSP Approached with ACS: Clustering-based Approaches vs. MinMax Formulation, EVOLVE-A Bridge between Probability, Set Oriented Numerics, and Evolutionary Computation VI, 210-223, Springer Cham, 2018.
  • R Necula, M Breaban, M Raschip. Tackling Dynamic Vehicle Routing Problem with Time Windows by means of ant colony system, IEEE Congress on Evolutionary Computation (CEC), 2480-2487, 2017.
  • R Necula, M Breaban, M Raschip. Tackling the bi-criteria facet of multiple traveling salesman problem with ant colony systems, IEEE 27th International Conference on Tools with Artificial Intelligence (ICTAI), 873-880, 2015.
  • S Timofeiov, M Marinca, C Bar, ME Breaban, V Drug, V Scripcariu. Conversion Rate to Resectability in Colorectal Cancer Liver Metastases: Need for Criteria Adapted to Current Therapy, Journal of Surgery [Jurnalul de chirurgie] 2015.
  • M. Breaban, A. Iftene. Dynamic Objective Sampling in Many-objective Optimization, Knowledge-Based and Intelligent Information & Engineering Systems 19th Annual Conference (KES 2015), Procedia Computer Science, Volume 60,Pages 178-187, 2015.
  • R. Necula, M. Breaban, M. Raschip. Performance Evaluation of Ant Colony Systems for the Single-Depot Multiple Traveling Salesman Problem, Hybrid Artificial Intelligent Systems (HAIS 2015), Springer International Publishing, pages 257-268, 2015.
  • H. Luchian, ME. Breaban, A. Bautu. On Meta-heuristics in Optimization and Data Analysis. Application to Geosciences, book chapter in Artificial Intelligent Approaches in Petroleum Geosciences, Springer International Publishing, pages 53-100, 2015.
  • S Timofeiov, ME Breabăn, V Drug, A Gervescu, I Huţanu. Extended Low Hartmann Operation with Total Mesorectal Excision-Optimal Surgical Treatment in Stage IV Mid and Upper Rectal Cancer, Journal of Surgery [Jurnalul de chirurgie] 10 (3), 117-120, 2014.
  • C. Cranganu, M. Breaban. Using support vector regression to estimate sonic log distributions: A case study from the Anadarko Basin, Oklahoma, Journal of Petroleum Science and Engineering, volume 103, pages 1-13, 2013.
  • C Serban, A Siriteanu, C Gheorghiu, A Iftene, L Alboaie, M Breaban. Combining Image Retrieval, Metadata Processing and Naive Bayes Classification at Plant Identification 2013, CLEF (Working Notes), 2013.
  • M. Breaban. Multiobjective Projection Pursuit for Semisupervised Feature Extraction. EVOApplications conference within EvoStar, LNCS 7835 Springer, pages 324-333, 2013.
    • abstract
      Abstract:
      The current paper presents a framework for linear feature extraction applicable in both unsupervised and supervised data analysis, as well as in their hybrid - the semi-supervised scenario. New features are extracted in a filter manner with a multi-modal genetic algorithm that optimizes simultaneously several projection indices. Experimental results show that the new algorithm is able to provide a compact and improved representation of the data set. The use of mixed labeled and unlabeled data under this scenario improves considerably the performance of constrained clustering algorithms such as constrained k-Means.
    • BibTeX
  • M. Breaban, H. Luchian, D. Simovici. A Genetic Clustering Algorithm by Monomial Projection Pursuit, International Symposium on Symbolic and Numeric Algorithms for Scientific Computing, 2012, to appear in post-proceedings IEEE 2013.
    • abstract
      Abstract:
      Aiming at exploratory analysis of data, this paper proposes a new method to identify interesting structures in data based on the projection pursuit methodology. Previous work reported in literature engage projection pursuit methods as a mean to visualize high-dimensional data or to identify linear combinations of attributes that reveal grouping tendencies or outliers. The framework of projection pursuit is generally formulated as an optimization problem aiming at finding projection axes that minimize/maximize a projection index. With regard to identifying interesting structure, the existing approaches suffer from obvious limitations: linear models are not able to catch more general structures in data like circular/curved clusters or any structure that is the result of a polynomial/nonlinear generative model. This paper extends linear projection pursuit to nonlinear projections while allowing at the same time for the preservation of the general methodology employed in the search of projections. In addition, an algorithmic framework based on multi-modal genetic algorithms is proposed in order to deal with the large search space and to allow for the use of non-differentiable projection indexes. Experiments conducted on synthetic data demonstrate the ability of the new approach to identify clusters of various shapes that otherwise are undetectable with linear projection pursuit or popular clustering methods like k-Means.
  • M. Breaban, H. Luchian Outlier Detection with Nonlinear Projection Pursuit. International Journal of Computers, Communications & Control Volume: 8 Issue: 1 Pages: 30-36 Published: FEB 2013
    • online paper
      Abstract:
      The current work proposes and investigates a new method to identify outliers in multivariate numerical data, driving its roots in projection pursuit. Projection pursuit is basically a method to deliver meaningful linear combinations of attributes. The novelty of our approach resides in introducing nonlinear combinations, able to model more complex interactions among attributes. The exponential increase of the search space with the increase of the polynomial degree is tackled with a genetic algorithm that performs monomial selection. Synthetic test cases highlight the benefits of the new approach over classical linear projection pursuit.
    • BibTeX
  • ME. Breaban, H. Luchian: PSO aided k-means clustering: introducing connectivity in k-means. GECCO 2011, pages 1227-1234, ACM 2011.
    • draft paper
      Abstract:
      Clustering is a fundamental and hence widely studied problem in data analysis. In a multi-objective perspective, this paper combines principles from two different clustering paradigms: the connectivity principle from density-based methods is integrated into the partitional clustering approach. The standard k-Means algorithm is hybridized with Particle Swarm Optimization. The new method (PSO-kMeans) beneffits from both a local and a global view on data and alleviates some drawbacks of the k-Means algorithm; thus, it is able to spot types of clusters which are otherwise difficult to obtain (elongated shapes, non-similar volumes). Our experimental results show that PSO-kMeans improves the performance of standard k-Means in all test cases and performs at least comparable to state-of-the-art methods in the worst case. PSO-kMeans is robust to outliers. This comes at a cost: the preprocessing step for nding the nearest neighbors for each data item is required, which increases the initial linear complexity of k-Means to quadratic complexity.
    • BibTeX
  • M. Breaban, H. Luchian. A unifying criterion for unsupervised clustering and feature selection. Pattern Recognition, 44(4), pages 854-865 (2011) ISSN 0031-3203.
    • abstract
      Abstract:
      Exploratory data analysis methods are essential for getting insight into data. Identifying the most important variables and detecting quasi-homogenous groups of data are problems of interest in this context. Solving such problems is a difficult task, mainly due to the unsupervised nature of the underlying learning process. Unsupervised feature selection and unsupervised clustering can be successfully approached as optimization problems by means of global optimization heuristics if an appropriate objective function is considered. This paper introduces an objective function capable of efficiently guiding the search for significant features and simultaneously for the respective optimal partitions. Experiments conducted on complex synthetic data suggest that the function we propose is unbiased with respect to both the number of clusters and the number of features.
    • BibTeX
  • M. Breaban, H. Luchian, DA. Simovici: Entropic-Genetic Clustering. EGC 2011, pages 71-76, 2011.
    • draft paper
      Abstract:
      This paper addresses the clustering problem given the similarity matrix of a dataset. We define two distinct criteria with the aim of simultaneously minimizing the cut size and obtaining balanced clusters. The first criterion minimizes the similarity between objects belonging to different clusters and is an objective generally met in clustering. The second criterion is formulated with the aid of generalized entropy. The trade-off between these two objectives is explored using a multi-objective genetic algorithm with enhanced operators.
    • BibTeX
  • M. Breaban. Evolving Ensembles of Feature Subsets towards Optimal Feature Selection for Unsupervised and Semi-supervised Clustering. IEA/AIE 2010, Part II, LNAI 6097, pages 67-76, Springer-Verlag, 2010
    • draft paper
      Abstract:
      The work in unsupervised learning centered on clustering has been extended with new paradigms to address the demands raised by real-world problems. In this regard, unsupervised feature selection has been proposed to remove noisy attributes that could mislead the clustering procedures. Additionally, semi-supervision has been integrated within existing paradigms because some background information usually exist in form of a reduced number of similarity/dissimilarity constraints. In this context, the current paper investigates a method to perform simultaneously feature selection and clustering. The benefits of a semi-supervised approach making use of reduced external information are highlighted against an unsupervised approach. The method makes use of an ensemble of near-optimal feature subsets delivered by a multi-modal genetic algorithm in order to quantify the relative importance of each feature to clustering.
    • BibTeX
  • M. Breaban. Optimized Ensembles for Clustering Noisy Data.Learning and Intelligent Optimization Workshop, in LNCS 6073, pages 220-223, Springer-Verlag, 2010
    • abstract
      Abstract:
      Clustering analysis is an important step towards getting insight into new data. Ensemble procedures have been designed in order to obtain improved partitions of a data set. Previous work in domain, mostly empirical, shows that accuracy and a limited diversity are mandatory features for successful ensemble construction. This paper presents a method which integrates unsupervised feature selection with ensemble clustering in order to deliver more accurate partitions. The efficiency of the method is studied on real data sets.
    • BibTeX
  • M. Ionita, M. Breaban, C. Croitoru.Evolutionary Computation in Constraint Satisfaction, book chapter in New Achievements in Evolutionary Computation", edt. Peter Korosec, INTECH Vienna, ISBN 978-953-307-053-7, 2010.
  • M. Breaban, H. Luchian. Unsupervised Feature Weighting with Multi Niche Crowding Genetic Algorithms. Genetic and Evolutionary Computation Conference, pages 1163-1170, ACM 2009
    • abstract
      Abstract:
      This paper is concerned with feature weighting/selection in the context of unsupervised clustering. Since different subspaces of the feature space may lead to different partitions of the data set, an efficient algorithm to tackle multi-modal environments is needed. In this context, the Multi-Niche Crowding Genetic Algorithm is used for searching relevant feature subsets. The proposed method is designed to deal with the inherent biases regarding the number of clusters and the number of features that appear in an unsupervised framework. The efficiency of the method is studied in the context of both feature weighting and feature selection using complex synthetic data.
    • BibTeX
  • M. Breaban, L. Alboaie, H. Luchian. Guiding Users within Trust Networks Using Swarm Algorithms. IEEE Congress on Evolutionary Computation,pages 1770-1777, IEEE Press, 2009
    • abstract
      Abstract:
      This paper is concerned with a problem in information organization and retrieval within Web communities. Most work in this domain is focused on reputation-based systems which exploit the experience gathered by previous users in order to evaluate resources at the community level. The current research focuses on a slightly different approach: a personalized evaluation system whose goal is to build a flexible and easy way to manage resources in a personalized manner. The functionality of such a model comes from local trust metrics which propagate the trust to a limited level into the system and, finally, lead to the appearance of minorities sharing some similar features/preferences. A modified PSO procedure is designed in order to analyze such a system and, in conjunction with a simple agglomerative clustering algorithm, identify homogenous groups of users.
    • BibTeX
  • M. Breaban, S. Luchian. Shaping up Clusters with PSO. In Proc. of 10th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing, Natural Computing and Applications Workshop, pages 532-537, IEEE press, 2008
    • abstract
      Abstract:
      This paper presents a method for enhancing the performance of current clustering algorithms; the method is based on Particle Swarm Optimization techniques. Namely, a pre-processing step aims at bringing closer objects which are likely to belong to the same cluster, while increasing the distance between objects likely to belong to different clusters. Experimental results show significantly improved performance for further clustering procedures especially when non-spherical clusters are involved.
  • M. Breaban, M. Ionita, C. Croitoru. A New PSO Approach to Constraint Satisfaction. In Proc. of IEEE Congress on Evolutionary Computation, pages 1948-1954,IEEE Press, september 2007
    • draft paper
      Abstract:
      Constraint satisfaction arises in many domains in different forms. Search and inference compete for solving constraint satisfaction problems (CSPs) but the most successful approaches are those which benefit from both techniques. Based on this idea, this article introduces a new scheme for solving the general Max-CSP problem. The new approach exploits the simplicity and efficiency of a modified Particle Swarm Optimization and the advantage of adaptable inference levels offered by the Mini-Bucket Elimination algorithm. Experiments conducted on binary CSPs using different levels of inference are illustrative for the inference/search trade-off. Comparative studies highlight the differences between our stochastic population-based method and the systematic search performed by a Branch and Bound algorithm.
    • BibTeX
  • M. Ionita, M. Breaban, C. Croitoru. A new scheme of using inference inside evolutionary computation techniques to solve CSPs. In Proc. of 8th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing, Natural Computing and Applications Workshop, pages 323-329, IEEE press, 2006.
  • M. Ionita, C. Croitoru, and M. Breaban. Incorporating inference into evolutionary algorithms for Max-CSP. In 3rd International Workshop on Hybrid Metaheuristics, LNCS 4030, pages 139-149, Springer-Verlag, 2006
  • M. Breaban, H. Luchian. PSO under an Adaptive Scheme. In Proc. of IEEE Congress on Evolutionary Computation, pages 1212- 1217 Vol. 2,IEEE Press, 2005.
    • draft paper
      Abstract:
      This paper presents an attempt to transform PSO into a self-adaptive algorithm based on specific swarm-inspired operators. New features are introduced: spatial expansion intended to overcome premature convergence (an algorithm called Improved PSO, IPSO) and auto-adaptation (an algorithm called Adaptive PSO, APSO). Experiments show that APSO and IPSO outperform the basic PSO on benchmark problems, proving their efficiency especially on multimodal functions.
    • BibTeX
Clustering:Evolutionary Approaches (PhD thesis)