Skip to main content

ESTADISTICA AVANZADA Y CIENCIA DE DATOS: INTERPRETANDO MODELOS CAJA-NEGRA Y ANALIZANDO CONJUNTOS DE DATOS GRANDES Y COMPLEJOS

PID2020-116294GB-I00

•
Funding agency name Agencia Estatal de Investigación
Funding agency acronym AEI
Program Programa Estatal de I+D+i Orientado a los Retos de la Sociedad
Call Proyectos de I+D+i (Generación de Conocimiento y Retos Investigación)
Call year 2020
State Plan 2017-2020 Plan Estatal de Investigación Científica y Técnica y de Innovación
Beneficiary centre UNIVERSITAT POLITECNICA DE CATALUNYA

Resultados relacionados

Resultados totales (Incluyendo duplicados): 11
Encontrada(s) 1 página(s)

UPCommons. Portal del coneixement obert de la UPC
Publication Artículo científico (JournalArticle)

Understanding complex predictive models with ghost variables

  • Delicado Useros, Pedro Francisco|||0000-0003-3933-4852
  • Peña Sanchez de Rivera, Daniel
The version of record of this article, first published in Test, is available online at Publisher’s website: http://dx.doi.org/10.1007/s11749-022-00826-x, Framed in the literature on Interpretable Machine Learning, we propose a new procedure to assign a measure of relevance to each explanatory variable in a complex predictive model. We assume that we have a training set to fit the model and a test set to check its out-of-sample performance. We propose to measure the individual relevance of each variable by comparing the predictions of the model in the test set with those obtained when the variable of interest is substituted (in the test set) by its ghost variable, defined as the prediction of this variable by using the rest of explanatory variables. In linear models it is shown that, on the one hand, the proposed measure gives similar results to leave-one-covariate-out (loco, with a lowest computational cost) and outperforms random permutations, and on the other hand, it is strongly related to the usual F-statistic measuring the significance of a variable. In nonlinear predictive models (as neural networks or random forests) the proposed measure shows the relevance of the variables in an efficient way, as shown by a simulation study comparing ghost variables with other alternative methods (including loco and random permutations, and also knockoff variables and estimated conditional distributions). Finally, we study the joint relevance of the variables by defining the relevance matrix as the covariance matrix of the vectors of effects on predictions when using every ghost variable. Our proposal is illustrated with simulated examples and the analysis of a large real data set., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publication Artículo científico (JournalArticle)

Multidimensional scaling for big data

  • Delicado Useros, Pedro Francisco|||0000-0003-3933-4852
  • Pachón García, Cristian|||0000-0001-9518-4874
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License. Open Access funding provided thanks to the CRUE-CSIC agreement with Springer Nature, We present a set of algorithms implementing multidimensional scaling (MDS) for large data sets. MDS is a family of dimensionality reduction techniques using a n x n distance matrix as input, where n is the number of individuals, and producing a low dimensional configuration: a n x r matrix with r << n. When n is large, MDS is unaffordable with classical MDS algorithms because their extremely large memory and time requirements. We compare six non-standard algorithms intended to overcome these difficulties. They are based on the central idea of partitioning the data set into small pieces, where classical MDS methods can work. Two of these algorithms are original proposals. In order to check the performance of the algorithms as well as to compare them, we have done a simulation study. Additionally, we have used the algorithms to obtain an MDS configuration for EMNIST: a real large data set with more than 800000 points. We conclude that all the algorithms are appropriate to use for obtaining an MDS configuration, but we recommend to use one of our proposals, since it is a fast algorithm with satisfactory statistical properties when working with big data. An R package implementing the algorithms has been created., This research was supported by the Spanish Research Agency (AEI) under project PID2020-116294GB-I00, by AGAUR under grants 2020 FI SDUR 306 and 2021 SGR 00613, and by UPC under AGRUPS-2022., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publication Capítulo o parte de un libro (BookPart)

Local constant likelihood estimation for beta distribution with time varying parameters

  • Acar Denizli, Nihan|||0000-0002-0012-8632
  • Delicado Useros, Pedro Francisco|||0000-0003-3933-4852
ThisresearchwassupportedbyprojectsPID2020-116294GB-I00andPID2023148158OB-I00 (founded by MICIU/AEI/10.13039/501100011033 and European Union NextGenerationEU/PRTR), by AGAUR under grant 2021 SGR 00613, and by UPC under AGRUPS-2022 and AGRUPS-2023. Nihan Acar-Denizli is a Serra-Húnter fellow., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publication Artículo científico (JournalArticle)

When are they coming? Understanding and forecasting the timeline of arrivals at the FC Barcelona stadium on match days

  • Serra Burriel, Feliu
  • Delicado Useros, Pedro Francisco|||0000-0003-3933-4852
  • Cucchietti, Fernando|||0000-0002-9027-1263
  • Graells Garrido, Eduardo
  • Gil Julian, Alex
  • Eguskiza Martínez, Imanol
Futbol Club Barcelona operates the largest stadium in Europe (with a seating capacity of almost one hundred thousand people) and manages recurring sports events. These are influenced by multiple conditions (time and day of the week, weather, adversary) and affect city dynamics—e.g., peak demand for related services like public transport and stores. We study fine grain audience entrances at the stadium segregated by visitor type and gate to gain insights and predict the arrival behavior of future games, with a direct impact on the organizational performance and productivity of the business. We can forecast the timeline of arrivals at gate level 72 h prior to kickoff, facilitating operational and organizational decision-making by anticipating potential agglomerations and audience behavior. Furthermore, we can identify patterns for different types of visitors and understand how relevant factors affect them. These findings directly impact commercial and business interests and can alter operational logistics, venue management, and safety., Open Access funding provided thanks to the CRUE-CSIC agreement with Springer Nature. This project has received funding from the European Union’s Horizon 2020 Research and Innovation Programme under the IoTwins Project (Grant agreement no. 857191). Pedro Delicado and Feliu Serra-Burriel would like to thank the Spanish Agencia Estatal de Investigación for the Grant PID2020-116294GB-I00., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publication Artículo científico (JournalArticle)

SurvLIMEpy: a Python package implementing SurvLIME

  • Pachón García, Cristian|||0000-0001-9518-4874
  • Hernández Pérez, Carlos|||0000-0002-0054-8573
  • Delicado Useros, Pedro Francisco|||0000-0003-3933-4852
  • Vilaplana Besler, Verónica|||0000-0001-6924-9961
© 2024 Elsevier. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/, In this paper we present SurvLIMEpy, an open-source Python package that implements the SurvLIME algorithm. This method allows to compute local feature importance for machine learning algorithms designed for modelling Survival Analysis data. The presented implementation uses a matrix-wise formulation, which allows to speed up the execution time. Additionally, SurvLIMEpy assists the user with visualisation tools to better understand the result of the algorithm. The package supports a wide variety of survival models, from the Cox Proportional Hazards Model to deep learning models such as DeepHit or DeepSurv. Two types of experiments are presented in this paper. First, by means of simulated data, we study the ability of the algorithm to capture the importance of the features. Second, we use three open source survival datasets together with a set of survival algorithms in order to demonstrate how SurvLIMEpy behaves when applied to different models., Peer Reviewed
O2, repositorio institucional de la UOC
Publication Artículo científico (JournalArticle)

Glycaemia fluctuations improvement in old-age prediabetic subjects consuming a quinoa-based diet: a pilot study

  • Díaz Rizzolo, Diana A
  • Acar-Denizli, Nihan
  • Kostov, Belchin
  • Roura, Elena
  • Sisó Almirall, Antoni
  • Delicado, Pedro
  • Gomis, Ramon
This study aimed to observe if quinoa could produce a benefit on postprandial glycemia that would result in less progression to type 2 diabetes (T2D). A cross-over design pilot clinical study with a nutritional intervention for 8 weeks was performed: 4 weeks on a regular diet (RD) and 4 weeks on a quinoa diet (QD). Nine subjects aged ≥65 years with prediabetes were monitored during the first 4 weeks of RD with daily dietary records and FreeStyle Libre®. Subsequently, participants started the QD, where quinoa and 100% quinoa-based products replaced foods rich in complex carbohydrates that they had consumed in the first 4 weeks of RD. The glycemic measurements recorded by the sensors were considered as functions of time, and the effects of nutrients consumed at the intended time period were analyzed by means of a function-on-scalar regression (fosr) model. With QD participants, decreased body weight (−1.6 kg, p = 0.008), BMI (−0.6 kg/m2p = 0.004) and waist circumference (−1.5 cm, p = 0.015) were observed. Nutrients intake changed during QD, namely, decreased carbohydrates (p = 0.004) and increased lipids (p = 0.004) and some amino acids (p < 0.05). The fosr model showed a reduction in postprandial glycemia in QD despite intrapersonal differences thanks to the joint action of different nutrients and the suppression of others consumed on a regular diet. We conclude that in an old age and high T2D-risk population, a diet rich in quinoa reduces postprandial glycemia and could be a promising T2D-preventive strategy.
Recercat. Dipósit de la Recerca de Catalunya
Publication Artículo científico (JournalArticle)

SurvLIMEpy: a Python package implementing SurvLIME

  • Pachón García, Cristian
  • Hernández Pérez, Carlos
  • Delicado Useros, Pedro Francisco
  • Vilaplana Besler, Verónica
© 2024 Elsevier. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/, In this paper we present SurvLIMEpy, an open-source Python package that implements the SurvLIME algorithm. This method allows to compute local feature importance for machine learning algorithms designed for modelling Survival Analysis data. The presented implementation uses a matrix-wise formulation, which allows to speed up the execution time. Additionally, SurvLIMEpy assists the user with visualisation tools to better understand the result of the algorithm. The package supports a wide variety of survival models, from the Cox Proportional Hazards Model to deep learning models such as DeepHit or DeepSurv. Two types of experiments are presented in this paper. First, by means of simulated data, we study the ability of the algorithm to capture the importance of the features. Second, we use three open source survival datasets together with a set of survival algorithms in order to demonstrate how SurvLIMEpy behaves when applied to different models., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publication

Local constant likelihood estimation for beta distribution with time varying parameters

  • Acar Denizli, Nihan
  • Delicado Useros, Pedro Francisco
ThisresearchwassupportedbyprojectsPID2020-116294GB-I00andPID2023148158OB-I00 (founded by MICIU/AEI/10.13039/501100011033 and European Union NextGenerationEU/PRTR), by AGAUR under grant 2021 SGR 00613, and by UPC under AGRUPS-2022 and AGRUPS-2023. Nihan Acar-Denizli is a Serra-Húnter fellow., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publication Artículo científico (JournalArticle)

Understanding complex predictive models with ghost variables

  • Delicado Useros, Pedro Francisco
  • Peña Sanchez de Rivera, Daniel
The version of record of this article, first published in Test, is available online at Publisher’s website: http://dx.doi.org/10.1007/s11749-022-00826-x, Framed in the literature on Interpretable Machine Learning, we propose a new procedure to assign a measure of relevance to each explanatory variable in a complex predictive model. We assume that we have a training set to fit the model and a test set to check its out-of-sample performance. We propose to measure the individual relevance of each variable by comparing the predictions of the model in the test set with those obtained when the variable of interest is substituted (in the test set) by its ghost variable, defined as the prediction of this variable by using the rest of explanatory variables. In linear models it is shown that, on the one hand, the proposed measure gives similar results to leave-one-covariate-out (loco, with a lowest computational cost) and outperforms random permutations, and on the other hand, it is strongly related to the usual F-statistic measuring the significance of a variable. In nonlinear predictive models (as neural networks or random forests) the proposed measure shows the relevance of the variables in an efficient way, as shown by a simulation study comparing ghost variables with other alternative methods (including loco and random permutations, and also knockoff variables and estimated conditional distributions). Finally, we study the joint relevance of the variables by defining the relevance matrix as the covariance matrix of the vectors of effects on predictions when using every ghost variable. Our proposal is illustrated with simulated examples and the analysis of a large real data set., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publication Artículo científico (JournalArticle)

When are they coming? Understanding and forecasting the timeline of arrivals at the FC Barcelona stadium on match days

  • Serra Burriel, Feliu
  • Delicado Useros, Pedro Francisco
  • Cucchietti, Fernando
  • Graells Garrido, Eduardo
  • Gil Julian, Alex
  • Eguskiza Martínez, Imanol
Futbol Club Barcelona operates the largest stadium in Europe (with a seating capacity of almost one hundred thousand people) and manages recurring sports events. These are influenced by multiple conditions (time and day of the week, weather, adversary) and affect city dynamics—e.g., peak demand for related services like public transport and stores. We study fine grain audience entrances at the stadium segregated by visitor type and gate to gain insights and predict the arrival behavior of future games, with a direct impact on the organizational performance and productivity of the business. We can forecast the timeline of arrivals at gate level 72 h prior to kickoff, facilitating operational and organizational decision-making by anticipating potential agglomerations and audience behavior. Furthermore, we can identify patterns for different types of visitors and understand how relevant factors affect them. These findings directly impact commercial and business interests and can alter operational logistics, venue management, and safety., Open Access funding provided thanks to the CRUE-CSIC agreement with Springer Nature. This project has received funding from the European Union’s Horizon 2020 Research and Innovation Programme under the IoTwins Project (Grant agreement no. 857191). Pedro Delicado and Feliu Serra-Burriel would like to thank the Spanish Agencia Estatal de Investigación for the Grant PID2020-116294GB-I00., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publication Artículo científico (JournalArticle)

Multidimensional scaling for big data

  • Delicado Useros, Pedro Francisco
  • Pachón García, Cristian
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License. Open Access funding provided thanks to the CRUE-CSIC agreement with Springer Nature, We present a set of algorithms implementing multidimensional scaling (MDS) for large data sets. MDS is a family of dimensionality reduction techniques using a n x n distance matrix as input, where n is the number of individuals, and producing a low dimensional configuration: a n x r matrix with r << n. When n is large, MDS is unaffordable with classical MDS algorithms because their extremely large memory and time requirements. We compare six non-standard algorithms intended to overcome these difficulties. They are based on the central idea of partitioning the data set into small pieces, where classical MDS methods can work. Two of these algorithms are original proposals. In order to check the performance of the algorithms as well as to compare them, we have done a simulation study. Additionally, we have used the algorithms to obtain an MDS configuration for EMNIST: a real large data set with more than 800000 points. We conclude that all the algorithms are appropriate to use for obtaining an MDS configuration, but we recommend to use one of our proposals, since it is a fast algorithm with satisfactory statistical properties when working with big data. An R package implementing the algorithms has been created., This research was supported by the Spanish Research Agency (AEI) under project PID2020-116294GB-I00, by AGAUR under grants 2020 FI SDUR 306 and 2021 SGR 00613, and by UPC under AGRUPS-2022., Peer Reviewed, Postprint (published version)