DESI - Artículos y ponencias con arbitraje
URI permanente para esta colecciónhttps://hdl.handle.net/11117/561
Examinar
Envíos recientes
Mostrando 1 - 20 de 281
Ítem Acceso Abierto Parallel k-Most Similar Neighbor Classifier for Mixed Data(Springer Nature, 2012-08) Sánchez-Díaz, Guillermo; Franco-Arcega, Anilu; Aguirre-Salado, Carlos; Piza-Dávila, Hugo I.; Morales-Manilla, Luis; Escobar-Franco, UrielThis paper presents a paralellization of the incremental algorithm inc-k-msn, for mixed data and similarity functions that do not satisfy metric properties. The algorithm presented is suitable for processing large data sets, because it only stores in main memory the k-most similar neighbors processed in step t, traversing only once the training data set. Several experiments with synthetic and real data are presented.Ítem Acceso Abierto A Fast Implementation for the Typical Testor Property Identification Based on an Accumulative Binary Tuple(Springer Nature, 2012-11) Sánchez-Díaz, Guillermo; Lazo-Cortés, Manuel; Piza-Dávila, Hugo I.In this paper, we introduce a fast implementation of the CT_EXT algorithm for testor property identification, that is based on an accumulative binary tuple. The fast implementation of the CT_EXT algorithm (one of the fastest algorithms reported), is designed to generate all the typical testers from a training matrix, requiring a reduced number of operations. Experimental results using this fast implementation and the comparison with other state-of-the-art algorithms that generate typical testers are presented.Ítem Embargo A parallel hill-climbing algorithm to generate a subset of irreducible testors(Springer Nature, 2014-11) Piza-Dávila, Hugo I.; Sánchez-Díaz, Guillermo; Aguirre-Salado, Carlos A.; Lazo-Cortés, ManuelThe generation of irreducible testors from a training matrix is an expensive computational process: all the algorithms reported have exponential complexity. However, for some problems there is no need to generate the entire set of irreducible testors, but only a subset of them. Several approaches have been developed for this purpose, ranging from Univariate Marginal Distribution to Genetic Algorithms. This paper introduces a parallel version of a Hill-Climbing Algorithm useful to find a subset of irreducible testors from a training matrix. This algorithm was selected because it has been one of the fastest algorithms reported in the state-of-the-art on irreducible testors. In order to efficiently store every different irreducible testor found, the algorithm incorporates a digital-search tree. Several experiments with synthetic and real data are presented in this work.Ítem Embargo A CUDA-based hill-climbing algorithm to find irreducible testors from a training matrix(Elsevier, 2017-05) Piza-Dávila, Hugo I.; Sánchez-Díaz, Guillermo; Lazo-Cortés, Manuel; Rizo-Domínguez, LuisIrreducible testors have been used to solve feature selection problems. All the exhaustive algorithms re- ported for the generation of irreducible testors have exponential complexity. However, several problems only require a portion of irreducible testors (only a subset of all). The hill-climbing algorithm is the lat- est approach that finds a subset of irreducible testors. So this paper introduces a parallel version of the hill-climbing algorithm which takes advantage of all the cores available in the graphics card because it has been developed on a CUDA platform. The proposed algorithm incorporates a novel mechanism that improves the exploration capability without adding any extra computation at the mutation step, thus increasing the rate of irreducible testors found. In addition, a Bloom filter is incorporated for efficient handling of duplicate irreducible testors. Several experiments with synthetic and real data, and a com- parison with other state-of-the-art algorithms are presented in this workÍtem Acceso Abierto Enhancing the Performance of YYC Algorithm Useful to Generate Irreducible Testors(World Scientific, 2017-07) Piza-Dávila, Hugo I.; Sánchez-Díaz, Guillermo; Lazo-Cortés, Manuel; Noyola-Medrano, CristinaIn pattern recognition, irreducible testors have been used for feature selection. A number of exhaustive algorithms that find irreducible testors have been reported in the literature. One of the latest and more efficient algorithms reported is YYC, an incremental algorithm that finds all the irreducible testors from a training matrix. Its efficiency relies on building a smaller number of feature combinations by finding compatible sets from the top of the matrix to the current row. Nevertheless, as the number of sets currently found grows, YYC execution becomes too slow. This work proposes two improvements of YYC algorithm, incorporated in a pre-processing phase; additionally, a parallel version is implemented. The paper presents some experimental results using synthetic and real data.Ítem Acceso Abierto An educational software for teaching database normalization(Wiley, 2017-06) Piza-Dávila, Hugo I.; Gutiérez-Preciado, Luis F.; Ortega-Guzmán, Víctor H.Database normalization is a key process for designing databases. However, it is one of the most complex topics for students in IT undergraduate programs. This work proposes a new software tool that supports students in the process of normalization. This learning tool allows the students to see step by step how to carry out each normal form up to the 3NF. This tool is based on data analysis of the input table as it is explained in class, and not on analyzing the scheme or initial dependencies defined by business rules. Thus, novel algorithms were developed to process the input table (user-defined) as well as the tables created during the normalization process. The software was tested in two Database courses, obtaining an important increase in the performance of students who used the tool.Ítem Embargo MapReduce approach to build network user profiles with top-k rankings for network security(IEEE, 2017-10) Parres-Peredo, Álvaro I.; Piza-Dávila, Hugo I.; Cervantes, FranciscoNetwork-user profiling has been used as security technique to detect unknown or malicious behaviors. Top-k rankings of reached services is a new technique for building user profiles. This technique requires to keep in memory all the traffic data during a period of time to build the rankings. However, a single user can produce gigabytes of network traffic data, which may result in low execution performance and out-of-memory errors. This work proposes a MapReduce approach that generates top-k rankings from huge network capture files.Ítem Embargo Building and Evaluating User Network Profiles for Cybersecurity Using Serverless Architecture(IEEE, 2019) Parres-Peredo, Álvaro I.; Piza-Dávila, Hugo I.; Cervantes, FranciscoServerless computing is a novel usage of cloud computing that offers programmers an environment to execute functions with no concerns on the infrastructure where the code runs. This work presents a serverless implementation of a technique used in cybersecurity, which focuses on the identification of unexpected behaviors from network users by means of TopK rankings. The implementation proposed leverages the potential of serverless computing by integrating the following services from Amazon Web Services: Lambda, DynamoDB, SQS and S3. The running time required to process user network traffic was dramatically reduced in comparison to other implementations of the same algorithm.Ítem Acceso Abierto Unexpected-Behavior Detection Using TopK Rankings for Cybersecurity(MDPI, 2019-10) Parres-Peredo, Álvaro I.; Piza-Dávila, Hugo I.; Cervantes, FranciscoAnomaly-based intrusion detection systems use profiles to characterize expected behavior of network users. Most of these systems characterize the entire network traffic within a single profile. This work proposes a user-level anomaly-based intrusion detection methodology using only the user’s network traffic. The proposed profile is a collection of TopK rankings of reached services by the user. To detect unexpected behaviors, the real-time traffic is organized into TopK rankings and compared to the profile using similarity measures. The experiments demonstrated that the proposed methodology was capable of detecting a particular kind of malware attack in all the users tested.Ítem Acceso Abierto An Algorithm for Computing Minimum-Length Irreducible Testors(IEEE, 2020-03) Piza-Dávila, Hugo I.; Guillermo Sanchez-Diaz; Lazo-Cortés, Manuel; Villalón-Turrubiates, Iván E.In pattern recognition, the elimination of unnecessary and/or redundant attributes is known as feature selection. Irreducible testors have been used to perform this task. An objective of the Minimum Description Length Principle (MDL) applied to feature selection in pattern recognition and data mining is to select the minimum number of attributes in a data set. Consequently, the MDL principle leads us to consider the subset of irreducible testors of minimum length. Some algorithms that nd the whole set of irreducible testors have been reported in the literature. However, none of these algorithms was designed to generate only minimum-length irreducible testors. In this paper, we propose the rst algorithm speci cally designed to calculate all minimum-length irreducible testors from a training sample. The paper presents some experimental results obtained using synthetic and real data in which the performance of the proposed algorithm is contrasted with other state-of-the-art algorithms that were adapted to generate only irreducible testers of minimum length.Ítem Acceso Abierto Genetic Algorithms for Piston and Tilt Detection by Using Young Patterns(MDPI, 2025-08) Piza-Dávila, Hugo I.; Salinas-Luna, Javier; Sánchez-Díaz, Guillermo; Chiu, Roger; Mora-González, MiguelWe present some numerical results on piston and tilt detection by using the Young experiment with Genetic Algorithms (GAs). We have simulated the cophasing of a flat surface by following the experimental setup and the mathematical model for Optical Path Difference (OPD) in the Young experiment to characterize piston and tip–tilt misalignment images in the order of a few nanometers, considering diffraction effects and random noise of 5%. Thus, the best fitness obtained by the genetic algorithm is considered as a determining factor to decide a complete error measurement because the proposed algorithm is capable of extracting the values of piston and tilt separately, regardless of which error is present or both. As a result, we have developed a study on piston detection from (0.001, 10) mm with a tilt present in the same pattern from (0, ./2) by using GAs embedded in a computational application.Ítem Acceso Abierto Modelling and Simulation Process of Extended Petri Nets with PNML and Matlab/Simulink(IJSIMM, 2023-06) Alcaraz-Mejía, M.; Parres-Peredo, Álvaro I.; Piza-Dávila, Hugo I.; Gutiérrez-Preciado, Luis F.In this work, we introduce a modelling and simulation process based on Matlab/Simulink for Petri nets that considers inhibitor arcs and priorities of transitions, herein named eXtended Petri nets, for modelling deterministic complex systems. The presented process comprises the method for modelling the system, the functions for the import to Matlab of all the data structures that define a Petri net with inhibitory arcs and priorities in transitions, and the functions to perform the simulation of the Petri net behaviour as a Simulink block. We present an exploratory case study about Rate Monotonic Scheduling for tasks with harmonic periods to show the complete modelling and simulation process. First, the scheduling system is modelled with Renew editor and saved as a PNML file. Then, this file is read and transformed into a Matlab data type. The produced data structures are the inputs to the proposed Simulink block, which performs the dynamic of the eXtended Petri net. Finally, the outputs of the simulation help validate the logical and temporal correctness of the scheduler model used as a case study. The concluding remarks section provides a link for downloading and testing the simulation process.Ítem Acceso Abierto A review of algorithms to computing irreducible testors applied to feature selection(Spring Nature, 2022-12) Sánchez-Díaz, Guillermo; Lazo‑Cortés, Manuel; Aguirre‑Salado, Carlos A.; Piza‑Dávila, Hugo I.; Garcia‑Contreras, Jorge P.Feature selection is an important task in the areas of pattern recognition and data mining. Various approaches to feature selection have been developed. In particular, this paper focuses on the algorithms for computing irreducible testors, which have been used to solve feature selection problems. The calculation of irreducible testors is an expensive computational process; the complexity of the algorithms to calculate the complete set of irreducible testors exponentially depends on the number of characteristics that describe the objects in the problem. To improve the execution time of these algorithms, different alternatives have been developed, such as parallel implementations, hardware-software implementation, rearrangement of the data, as well as heuristics to generate just an irreducible testor or a subset of the entire set of irreducible testors, among other strategies. This paper presents a review of the literature on irreducible testors, with the aim of providing a guide for researchers working in the areas of pattern recognition and data mining, interested in feature selection, using heterogeneous data and possibly missing dataÍtem Acceso Abierto CdbNorm: An Efficient Library for Automatic Database Normalization(The Science and Information Organization, 2026-03) Piza-Dávila, Hugo I.; Gutiérrez-Preciado, Luis F.; Ortega-Guzmán, Víctor H.; Alcaraz-Mejía, MildrethThis study introduces CdbNorm, a library that provides efficient implementations of the first three normal forms of relational database normalization. CdbNorm makes it quick and straightforward for a data analyst to divide a large dataset into smaller tables free from database anomalies (insert, update, and delete) and duplicate data. This study describes each of the steps of our normalization algorithm, which includes the discovery of functional dependencies and the population of output normalized datasets. We evaluate the accuracy and efficiency of our algorithm with databases introduced in prior papers and with large datasets available online.Ítem Acceso Abierto Space Mapping with Parameter Extraction Based on the Kullback-Leibler Distance Illustrated with Full-Wave EM and Equivalent Circuit Models for Microstrip Filters(IEEE, 2025-07-21) Loera-Díaz, Roberto; Rayas-Sánchez, José E.Space mapping (SM) techniques are commonly used for optimizing highly accurate models that require a large computational effort, known as fine models, by exploiting simplified physics-based models that are computationally fast but not accurate enough, known as coarse models. Most SM formulations require solving a parameter extraction (PE) sub-problem at each iteration. Typically, SM algorithms use classical l-th norms for the PE objective function. In this paper, we first apply a PE formulation based on the Kullback-Leibler (K-L) distance to microstrip filters using their full-wave electromagnetic (EM) responses as targets, performing a rigorous numerical comparison against PE using classical norms. We subsequently propose, for the first time, a Broyden-based input space mapping algorithm using the K-L distance as objective function for the PE sub-problem. We apply SM design optimization to several examples, beginning with a classical synthetic test example and following with some microstrip filters using their full-wave EM representation as fine models, and their equivalent distributed circuit as coarse models. A rigorous numerical comparison is also performed between classical l-th norms and the K-L formulation for PE within the corresponding SM design optimizations. Our results indicate that SM with PE using the K-L formulation outperforms that one obtained by using the classical l-th norm PE formulations within SM.Ítem Acceso Abierto Cognitive Broyden-based Input Space Mapping for Design Optimization(IEEE, 2025-04-28) Rayas-Sánchez, José E.Cognition-driven design of RF and microwave circuits is an emerging and promising approach to efficient design optimization of computationally expensive fine models. Existing techniques for cognition-driven design have been developed for optimizing microwave filters without exploiting traditional coarse model representations, e.g., equivalent circuits. Instead, intermediate feature-space parameters have been used to establish other types of mappings in the design process. In this paper, a cognitive space mapping (SM) technique that fully exploits traditional coarse models is proposed for the first time. The proposed cognitive SM approach exploits a previous cognition-driven parameter extraction (PE) formulation at each SM iteration. This cognitive SM technique follows an algorithmic structure that is an extension of that one used by the Broyden-based input SM, better known as aggressive space mapping (ASM). A synthetic benchmark example illustrates the performance improvement of the proposed cognitive SM versus ASM.Ítem Acceso Abierto Cognitive Parameter Extraction for Space Mapping Design Optimization(IEEE, 2025) Rayas-Sánchez, José E.Parameter extraction (PE) is a key subproblem of space mapping (SM) design optimization. It consists of a local alignment between the coarse and fine models at each SM iteration. In this work, cognition-driven PE is proposed for SM. In contrast to classical PE, where the full fine model responses are used as targets, the proposed cognitive PE focuses on key features of the fine model response selected from an engineering perspective. It is demonstrated that the proposed cognitive PE approach: 1) yields more accurate extracted parameters regardless of the type of PE objective function employed; and 2) achieves a more meaningful matching to the fine model target response and with less variability. To proof this with independence of the optimization method employed for PE, plots of the PE objective functions are presented over large regions of the coarse model design space. Two synthetic examples are used to support these findings.Ítem Acceso Abierto AI-Enhanced Post-Silicon Validation: Tackling Signal Integrity Issues in High-Speed Interconnects(IEEE, 2025-01-23) Rangel-Patiño, Francisco; Vega-Ochoa, Édgar A.; Viveros-Wacher, Andrés; Onsongo, Daudi; Rayas-Sánchez, José E.Semiconductor technology advances, coupled with the demand for higher data rates and bandwidth, has led to significant signal integrity issues such as attenuation, crosstalk, jitter, noise, EM susceptibility, etc. Traditional post-silicon validation methods for high-performance computer platforms, which rely heavily on manual inspection and rule-based heuristics, are increasingly inadequate for addressing these complexities. In this paper, we review and highlight the application of artificial intelligence (AI) approaches and machine learning (ML) techniques to automate post-silicon validation and enhance the detection and diagnosis of signal integrity issues in high-speed computer interfaces. Through a series of case studies, we demonstrate the efficacy of various AI techniques, including artificial neural networks (ANNs), surrogate modeling, and unsupervised learning, in optimizing settings and improving the efficiency of post-silicon validation. These techniques significantly reduce the number of required measurements, enhance accuracy, and provide scalable and flexible solutions for modern post-silicon physical layer validation and tuning processes.Ítem Acceso Abierto A Rigorous Numerical Comparison for Parameter Extraction Using Classical and Kullback-Leibler Formulations Illustrated with Full-Wave EM and Equivalent Circuit Models for Microstrip Filters(IEEE, 2025-01) Loera-Díaz, Roberto; Rayas-Sánchez, José E.; Villa-Loustaunau, Enrique R.Numerical circuit parameter extraction (PE) is a key sub-process of space mapping (SM), which is used to efficiently optimize full-wave EM responses of microwave structures exploiting faster but inaccurate physics-based auxiliary models. Any improvement in PE has a positive impact on SM design optimization. In this paper, we apply for the first time a PE formulation based on the Kullback-Leibler distance to microstrip filters using their full-wave EM responses as targets. We perform a rigorous numerical comparison of PE based on the K-L formulation against PE using classical norms. Our results confirm a better PE performance using the Kullback-Leibler formulation than those obtained with traditional PE formulations.Ítem Acceso Abierto Optimizing the Full Mixed-Mode Performance of a Differential Microstrip Interconnect with a Right-Angle Bend by using Two Symmetrical Bumps(IEEE, 2025-01) Villa-Loustaunau, Enrique R.; Rayas-Sánchez, José E.; Loera-Díaz, Roberto; Rangel-Patiño, FranciscoDifferential interconnects are widely used for high-speed serial data transmission in modern high-performance computer platforms. Differential signaling handle noise better than single ended signaling. However, physical asymmetries and discontinuities in differential links can cause that a portion of the differential energy is converted into common mode (CM) energy, which is perceived as noise at the receiver. This mode conversion in differential interconnects leads to electromagnetic (EM) interference and EM susceptibility, limiting high data rates. In this paper, a microstrip differential interconnect with a severe discontinuity, a right-angle bend, is optimally compensated by using two rectangular length-match bumps. Our formulation allows the efficient optimization of the full set of mixed-mode (MM) S-parameters of the differential interconnect. It uses a smart combination of pattern search and Nelder-Mead to optimize the MM performance considering several starting points. The interconnect MM performance before and after optimization is shown, confirming a very significant performance improvement.