Stéphane Bonnevay

A pretopological approach for structural analysis

Abstract The aim of this paper is to present a methodological approach for problems encountered in structural analysis. This approach is based upon the pretopological concepts of pseudoclosure and minimal closed subsets. The advantage of this approach is that it provides a framework which is general enough to model and formulate different types of connections that exist between the elements of a population. In addition, it has enabled us to develop a new structural analysis algorithm. An explanation of the definitions and properties of the pretopological concepts applied in this work is first shown and illustrated in sample settings. The structural analysis algorithm is then described and the results obtained in an economic study of the impact of geographic proximity on scientific collaborations are presented.

Time-to-event analysis with artificial neural networks: An integrated analytical and rule-based study for breast cancer

This paper presents an analysis of censored survival data for breast cancer specific mortality and disease-free survival. There are three stages to the process, namely time-to-event modelling, risk stratification by predicted outcome and model interpretation using rule extraction. Model selection was carried out using the benchmark linear model, Cox regression but risk staging was derived with Cox regression and with Partial Logistic Regression Artificial Neural Networks regularised with Automatic Relevance Determination (PLANN-ARD). This analysis compares the two approaches showing the benefit of using the neural network framework especially for patients at high risk. The neural network model also has results in a smooth model of the hazard without the need for limiting assumptions of proportionality. The model predictions were verified using out-of-sample testing with the mortality model also compared with two other prognostic models called TNG and the NPI rule model. Further verification was carried out by comparing marginal estimates of the predicted and actual cumulative hazards. It was also observed that doctors seem to treat mortality and disease-free models as equivalent, so a further analysis was performed to observe if this was the case. The analysis was extended with automatic rule generation using Orthogonal Search Rule Extraction (OSRE). This methodology translates analytical risk scores into the language of the clinical domain, enabling direct validation of the operation of the Cox or neural network model. This paper extends the existing OSRE methodology to data sets that include continuous-valued variables.

Comparison of artificial neural network with logistic regression as classification models for variable selection for prediction of breast cancer patient outcomes

The aim of this study was to compare multilayer perceptron neural networks (NNs) with standard logistic regression (LR) to identify key covariates impacting on mortality from cancer causes, disease-free survival (DFS), and disease recurrence using Area Under Receiver-Operating Characteristics (AUROC) in breast cancer patients. From 1996 to 2004, 2,535 patients diagnosed with primary breast cancer entered into the study at a single French centre, where they received standard treatment. For specific mortality as well as DFS analysis, the ROC curves were greater with the NN models compared to LR model with better sensitivity and specificity. Four predictive factors were retained by both approaches for mortality: clinical size stage, Scarff Bloom Richardson grade, number of invaded nodes, and progesterone receptor. The results enhanced the relevance of the use of NN models in predictive analysis in oncology, which appeared to be more accurate in prediction in this French breast cancer cohort.

Hybrid Metaheuristics based on MOEA/D for 0/1 multiobjective knapsack problems: A comparative study

Hybrid Metaheuristics aim to incorporate and combine different metaheuristics with each other to enhance the search capabilities. It can improve both of intensification and diversification toward the preferred solutions and concentrates the search efforts to investigate the promising regions in the search space. In this paper, a comparative study was developed to study the effect of the hybridization of different metaheuris- tics within MOEA/D framework. We study four proposals of hybridization, the first proposal is to combine adaptive discrete differential evolution operator with MOEA/D. The second one is to combine the path-Relinking operator with MOEA/D. the third and the fourth proposals combine both of them in MOEA/D. The comparative study uses a set of MOKSP instances commonly used in the literature to investigate the hybridization effects as well as a set of quality assessment indicators. The experimental results indicate that the proposals are highly competitive for most test instances and can be considered as viable alternatives.

Breast Cancer Predictions by Neural Networks Analysis: a Comparison with Logistic Regression

This paper presents an exploratory fixed time study to identify the most significant covariates as a precursor to a longitudinal study of specific mortality, disease free survival and disease recurrences. The data comprise consecutive patients diagnosed with primary breast cancer and entered into the study from 1996 at a single French clinical center, Centre Leon Berard, based in Lyon, where they received standard treatment. The methodology was to compare and contrast multi-layer perceptron neural networks (NN) with logistic regression (LR), to identify key covariates and their interactions and to compare the selected variables with those routinely used in clinical severity of illness indices for breast cancer. The logistic regression in this work was chosen as an accepted standard for prediction by biostatisticians in order to evaluate the neural network. Only covariates available at the time of diagnosis and immediately following surgery were used. We used for comparison classification performance indices: AUROC (AREA Under Receiver-Operating Characteristics) curves, sensitivity, specificity, accuracy and positive predictive value for the two following events of interest: specific mortality and disease free survival.

A hybrid evolutionary metaheuristics (HEMH) applied on 0/1 multiobjective knapsack problems

Handling Multiobjective Optimization Problems (MOOP) using Hybrid Metaheuristics represents a promising and interest area of research. In this paper, a Hybrid Evolutionary Metaheuristics (HEMH) is presented. It combines different metaheuristics integrated with each other to enhance the search capabilities. It improves both of intensification and diversification toward the preferred solutions and concentrates the search efforts to investigate the promising regions in the search space. In the proposed HEMH, the search process is divided into two phases. In the first one, the DM-GRASP is applied to obtain an initial set of high quality solutions dispersed along the Pareto front. Then, the search efforts are intensified on the promising regions around these solutions through the second phase. The greedy randomized path-relinking with local search or reproduction operators are applied to improve the quality and to guide the search to explore the non discovered regions in the search space. The two phases are combined with a suitable evolutionary framework supporting the integration and cooperation. Moreover, the efficient solutions explored over the search are collected in an external archive. The HEMH is verified and tested against some of the state of the art MOEAs using a set of MOKSP instances commonly used in the literature. The experimental results indicate that the HEMH is highly competitive and can be considered as a viable alternative.

Temporal Multinomial Mixture for Instance-Oriented Evolutionary Clustering

Evolutionary clustering aims at capturing the temporal evolution of clusters. This issue is particularly important in the context of social media data that are naturally temporally driven. In this paper, we propose a new probabilistic model-based evolutionary clustering technique. The Temporal Multinomial Mixture (TMM) is an extension of classical mixture model that optimizes feature co-occurrences in the trade-off with temporal smoothness. Our model is evaluated for two recent case studies on opinion aggregation over time. We compare four different probabilistic clustering models and we show the superiority of our proposal in the task of instance-oriented clustering.

Data Analysis Based on Minimal Closed Subsets

The aim of this paper is to provide a framework which enables us to treat structural analysis problems. This framework is based on pretopological theory. We apply the concepts of pseudoclosure and minimal closed subsets to bring out the structural information. In order to illustrate our method, an application to co-authorships of publications between French geographical areas is displayed.

Simultaneous Clustering and Model Selection for Multinomial Distribution: A Comparative Study

In this paper, we study different discrete data clustering methods, which use the Model-Based Clustering (MBC) framework with the Multinomial distribution. Our study comprises several relevant issues, such as initialization, model estimation and model selection. Additionally, we propose a novel MBC method by efficiently combining the partitional and hierarchical clustering techniques. We conduct experiments on both synthetic and real data and evaluate the methods using accuracy, stability and computation time. Our study identifies appropriate strategies to be used for discrete data analysis with the MBC methods. Moreover, our proposed method is very competitive w.r.t. clustering accuracy and better w.r.t. stability and computation time.

A hybrid evolutionary approach with search strategy adaptation for mutiobjective optimization

Hybrid evolutionary algorithms have been successfully applied to solve numerous multiobjective optimization problems (MOP). In this paper, a new hybrid evolutionary approach based on search strategy adaptation (HESSA) is presented. In HESSA, the search process is carried out through adopting a pool of different search strategies, each of which has a specified success ratio. A new offspring is generated using a randomly selected strategy. Then, according to the success of the generated offspring to update the population or the archive, the success ratio of the selected strategy is adapted. This provides the ability for HESSA to adopt the appropriate search strategy according to the problem on hand. Furthermore, the cooperation among different strategies leads to improve the exploration and the exploitation of the search space. The proposed pool is combined to a suitable evolutionary framework for supporting the integration and cooperation. Moreover, the efficient solutions explored over the search are collected in an external repository to be used as global guides. The proposed HESSA is verified against some of the state of the art MOEAs using a set of test problems commonly used in the literature. The experimental results indicate that HESSA is highly competitive and can be considered as a viable alternative.