Search results for " optimization."
showing 10 items of 2333 documents
Energy Efficiency Optimization for Multi-cell Massive MIMO : Centralized and Distributed Power Allocation Algorithms
2021
This paper investigates the energy efficiency (EE) optimization in downlink multi-cell massive multiple-input multiple-output (MIMO). In our research, the statistical channel state information (CSI) is exploited to reduce the signaling overhead. To maximize the minimum EE among the neighbouring cells, we design the transmit covariance matrices for each base station (BS). Specifically, optimization schemes for this max-min EE problem are developed, in the centralized and distributed ways, respectively. To obtain the transmit covariance matrices, we first find out the closed-form optimal transmit eigenmatrices for the BS in each cell, and convert the original transmit covariance matrices desi…
From optimization to algorithmic differentiation: a graph detour
2021
This manuscript highlights the work of the author since he was nominated as "Chargé de Recherche" (research scientist) at Centre national de la recherche scientifique (CNRS) in 2015. In particular, the author shows a thematic and chronological evolution of his research interests:- The first part, following his post-doctoral work, is concerned with the development of new algorithms for non-smooth optimization.- The second part is the heart of his research in 2020. It is focused on the analysis of machine learning methods for graph (signal) processing.- Finally, the third and last part, oriented towards the future, is concerned with (automatic or not) differentiation of algorithms for learnin…
Optimisation des requêtes de similarité dans les espaces métriques répondant aux besoins des usagers
2012
The complexity of data stored in large databases has increased at very fast paces. Hence, operations more elaborated than traditional queries are essential in order to extract all required information from the database. Therefore, the interest of the database community in similarity search has increased significantly. Two of the well-known types of similarity search are the Range (Rq) and the k-Nearest Neighbor (kNNq) queries, which, as any of the traditional ones, can be sped up by indexing structures of the Database Management System (DBMS). Another way of speeding up queries is to perform query optimization. In this process, metrics about data are collected and employed to adjust the par…
Computing the Arrangement of Circles on a Sphere, with Applications in Structural Biology
2009
International audience; Balls and spheres are the simplest modeling primitives after affine ones, which accounts for their ubiquitousness in Computer Science and Applied Mathematics. Amongst the many applications, we may cite their prevalence when it comes to modeling our ambient 3D space, or to handle molecular shapes using Van der Waals models. If most of the applications developed so far are based upon simple geometric tests between balls, in particular the intersection test, a number of applications would obviously benefit from finer pieces of information. Consider a sphere $S_0$ and a list of circles on it, each such circle stemming from the intersection between $S_0$ and another spher…
Simplification of Models
2016
In practical applications the “complete” model, i.e., a model that contains all features that the experts in the application domain consider important, is often quite complicated and difficult to analyse mathematically. A straightforward numerical realization is often costly and may give very little qualitative understanding of the situation. It is therefore important to study if the model can be systematically simplified in order to enhance a qualitative analysis/understanding.
Further results on H<inf>&#x221E;</inf> control of switched linear time-delay systems
2012
In this note, we study the problems of stability analysis and H ∞ controller synthesis of discrete-time switched systems with time-varying delay. The system under consideration is firstly transformed into an interconnection system. Based on the system transformation and the scaled small gain theorem, the asymptotic stability of the original system is examined via the version of the bounded realness of the transformed forward system. The aim of the proposed approach is to reduce conservatism, which is made possible by a precise approximation of the time-varying delay and the input-output approach. The proposed stability condition is demonstrated to be much less conservative than most existin…
Fuzzy filter design for discrete-time delayed systems with distributed probabilistic sensor faults
2013
In this paper, the problem of distributed fuzzy filter design has been solved for T-S fuzzy systems with time-varying delays and multiple probabilistic packet losses. Our attention is paid to designing the distributed fuzzy filters to guarantee the filtering error dynamic system to be mean-square asymptotically stable with an average ℋ∞ performance. Sufficient conditions for the obtained filtering error dynamic system are proposed by applying a comparison model and the scaled small gain theorem. Based on the measurements and estimates of the system states for each sensor and its neighbors, the solution of the parameters of the distributed fuzzy filters is characterized in terms of the feasi…
SIOPRED performance in a Forecasting Blind Competition
2012
In this paper we present the results obtained by applying our automatic forecasting support system, named SIOPRED, over a data set of time series in a Forecasting Blind Competition. In order to apply our procedure for providing point forecasts it has been necessary to develop an interactive strategy for the choice of the suitable length of the seasonal cycle and the seasonality form for a generalized exponential smoothing method, which have been obtained using SIOPRED. For the choice of those essential characteristics of forecasting methods, also a certain multi-objective formulation which minimizes several measures of fitting is used. Once these specifications are established, the model pa…
Soft Computing Techniques for Portfolio Selection: Combining SRI with Mean-Variance Goals
2014
A fuzzy portfolio selection model is presented incorporating a socially responsible goal without discarding a priori financially good portfolios or weakening a priori the financial goals. Hence, the optimal portfolios it provides could be either efficient from the strictly financial point of view or non-efficient if leaving the efficient frontier substantially improves the degree of social responsibility. The model can be used to direct heuristic procedures in order to select a reduced number of various alternatives from which the investor can directly make a final decision.
Network-Based Computational Techniques to Determine the Risk Drivers of Bank Failures During a Systemic Banking Crisis
2018
This paper employs a computational model of solvency and liquidity contagion assessing the vulnerability of banks to systemic risk. We find that the main risk drivers relate to the financial connections a bank has and the market concentration, apart from the size of the bank triggering the contagion, while balance sheets play only a minor role. We also find that market concentration might facilitate banks to withstand liquidity shocks better while exposing them to larger solvency chocks. Our results are validated through an out-of-sample forecasting that shows that both type I and type II prediction errors are reduced if we include network characteristics in our prediction model.