|   | 
Details
   web
Records
Author Guevara, E.; Babonneau, F.; Homem-de-Mello, T.; Moret, S.
Title A machine learning and distributionally robust optimization framework for strategic energy planning under uncertainty Type Journal Article
Year 2020 Publication Applied Energy Abbreviated Journal Appl. Energy
Volume 271 Issue Pages 18 pp
Keywords Strategic energy planning; Electricity generation; Uncertainty; Distributionally robust optimization; Machine learning
Abstract This paper investigates how the choice of stochastic approaches and distribution assumptions impacts strategic investment decisions in energy planning problems. We formulate a two-stage stochastic programming model assuming different distributions for the input parameters and show that there is significant discrepancy among the associated stochastic solutions and other robust solutions published in the literature. To remedy this sensitivity issue, we propose a combined machine learning and distributionally robust optimization (DRO) approach which produces more robust and stable strategic investment decisions with respect to uncertainty assumptions. DRO is applied to deal with ambiguous probability distributions and Machine Learning is used to restrict the DRO model to a subset of important uncertain parameters ensuring computational tractability. Finally, we perform an out-of-sample simulation process to evaluate solutions performances. The Swiss energy system is used as a case study all along the paper to validate the approach.
Address [Guevara, Esnil] Univ Adolfo Ibanez, PhD Program Ind Engn & Operat Res, Santiago, Chile, Email: frederic.babonneau@uai.cl
Corporate Author Thesis
Publisher Elsevier Sci Ltd Place of Publication Editor
Language English Summary Language Original Title
Series Editor Series Title Abbreviated Series Title
Series Volume Series Issue Edition
ISSN 0306-2619 ISBN Medium
Area Expedition Conference
Notes WOS:000540436500003 Approved no
Call Number UAI @ eduardo.moreno @ Serial 1188
Permanent link to this record
 

 
Author Arevalo-Ramirez, T.; Villacres, J.; Fuentes, A.; Reszka, P.; Cheein, F.A.A.
Title Moisture content estimation of Pinus radiata and Eucalyptus globulus from reconstructed leaf reflectance in the SWIR region Type Journal Article
Year 2020 Publication Biosystems Engineering Abbreviated Journal Biosyst. Eng.
Volume 193 Issue Pages 187-205
Keywords SWIR reconstruction; Machine learning; Spectral indices; Fuel moisture content; Equivalent water thickness
Abstract Valparaiso, a central-southern region in Chile, has one of the highest rates of wildfire occurrence in the country. The constant threat of fires is mainly due to its highly flammable forest plantation, composed of 97.5% Pinus radiata and Eucalyptus globulus. Fuel moisture content is one of the most relevant parameters for studying fire spreading and risk, and can be estimated from the reflectance of leaves in the short wave infra-red (SWIR) range, not easily available in most vision-based sensors. Therefore, this work addresses the problem of estimating the water content of leaves from the two previously mentioned species, without any knowledge of their spectrum in the SWIR band. To this end, and for validation purposes, the reflectance of 90 leaves per species, at five dehydration stages, were taken between 350 nm and 2500 nm (full spectrum). Then, two machine-learning regressors were trained with 70% of the data set to determine the unknown reflectance, in the range 1000 nm-2500 nm. Results were validated with the remaining 30% of the data, achieving a root mean square error less than 9% in the spectrum estimation, and an error of 10% in spectral indices related to water content estimation. (C) 2020 IAgrE. Published by Elsevier Ltd. All rights reserved.
Address [Arevalo-Ramirez, Tito; Villacres, Juan; Auat Cheein, Fernando A.] Univ Tecn Fedrico Santa Maria, Dept Ingn Elect, Valparaiso, Chile, Email: fernando.auat@usm.cl
Corporate Author Thesis
Publisher Academic Press Inc Elsevier Science Place of Publication Editor
Language English Summary Language Original Title
Series Editor Series Title Abbreviated Series Title
Series Volume Series Issue Edition
ISSN 1537-5110 ISBN Medium
Area Expedition Conference
Notes WOS:000526114500016 Approved no
Call Number UAI @ eduardo.moreno @ Serial 1150
Permanent link to this record
 

 
Author Hughes, S.; Moreno, S.; Yushimito, W.F.; Huerta-Canepa, G.
Title Evaluation of machine learning methodologies to predict stop delivery times from GPS data Type Journal Article
Year 2019 Publication Transportation Research Part C-Emerging Technologies Abbreviated Journal Transp. Res. Pt. C-Emerg. Technol.
Volume 109 Issue Pages 289-304
Keywords Machine learning; Stop delivery time; Classification; Regression; Hazard duration; GPS
Abstract In last mile distribution, logistics companies typically arrange and plan their routes based on broad estimates of stop delivery times (i.e., the time spent at each stop to deliver goods to final receivers). If these estimates are not accurate, the level of service is degraded, as the promised time window may not be satisfied. The purpose of this work is to assess the feasibility of machine learning techniques to predict stop delivery times. This is done by testing a wide range of machine learning techniques (including different types of ensembles) to (1) predict the stop delivery time and (2) to determine whether the total stop delivery time will exceed a predefined time threshold (classification approach). For the assessment, all models are trained using information generated from GPS data collected in Medellin, Colombia and compared to hazard duration models. The results are threefold. First, the assessment shows that regression-based machine learning approaches are not better than conventional hazard duration models concerning absolute errors of the prediction of the stop delivery times. Second, when the problem is addressed by a classification scheme in which the prediction is aimed to guide whether a stop time will exceed a predefined time, a basic K-nearest-neighbor model outperforms hazard duration models and other machine learning techniques both in accuracy and F-1 score (harmonic mean between precision and recall). Third, the prediction of the exact duration can be improved by combining the classifiers and prediction models or hazard duration models in a two level scheme (first classification then prediction). However, the improvement depends largely on the correct classification (first level).
Address [Hughes, Sebastian; Moreno, Sebastian; Yushimito, Wilfredo F.; Huerta-Canepa, Gonzalo] Univ Adolfo Ibanez, Fac Engn & Sci, Vina Del Mar, Chile, Email: shughes@alumnos.uai.cl;
Corporate Author Thesis
Publisher Pergamon-Elsevier Science Ltd Place of Publication Editor
Language English Summary Language Original Title
Series Editor Series Title Abbreviated Series Title
Series Volume Series Issue Edition
ISSN 0968-090x ISBN Medium
Area Expedition Conference
Notes WOS:000504780800016 Approved no
Call Number UAI @ eduardo.moreno @ Serial 1082
Permanent link to this record
 

 
Author Alejo, L.; Atkinson, J.; Guzman-Fierro, V.; Roeckel, M.
Title Effluent composition prediction of a two-stage anaerobic digestion process: machine learning and stoichiometry techniques Type Journal Article
Year 2018 Publication Environmental Science And Pollution Research Abbreviated Journal Environ. Sci. Pollut. Res.
Volume 25 Issue 21 Pages 21149-21163
Keywords Anaerobic digestion; Protein degradation; Machine learning; Prediction methods; Support vector machines
Abstract Computational self-adapting methods (Support Vector Machines, SVM) are compared with an analytical method in effluent composition prediction of a two-stage anaerobic digestion (AD) process. Experimental data for the AD of poultry manure were used. The analytical method considers the protein as the only source of ammonia production in AD after degradation. Total ammonia nitrogen (TAN), total solids (TS), chemical oxygen demand (COD), and total volatile solids (TVS) were measured in the influent and effluent of the process. The TAN concentration in the effluent was predicted, this being the most inhibiting and polluting compound in AD. Despite the limited data available, the SVM-based model outperformed the analytical method for the TAN prediction, achieving a relative average error of 15.2% against 43% for the analytical method. Moreover, SVM showed higher prediction accuracy in comparison with Artificial Neural Networks. This result reveals the future promise of SVM for prediction in non-linear and dynamic AD processes.
Address [Alejo, Luz; Guzman-Fierro, Victor; Roeckel, Marlene] Univ Concepcion, Dept Ingn Quim, Victor Lamas 1290,Casilla 160-C Correo 3., Concepcion, Chile, Email: john.atkinson@uai.cl;
Corporate Author Thesis
Publisher Springer Heidelberg Place of Publication Editor
Language English Summary Language Original Title
Series Editor Series Title Abbreviated Series Title
Series Volume Series Issue Edition
ISSN 0944-1344 ISBN Medium
Area Expedition Conference
Notes WOS:000438830900080 Approved no
Call Number UAI @ eduardo.moreno @ Serial 890
Permanent link to this record
 

 
Author Pham, D.T.; Ruz, G.A.
Title Unsupervised training of Bayesian networks for data clustering Type Journal Article
Year 2009 Publication Proceedings Of The Royal Society A-Mathematical Physical And Engineering Sciences Abbreviated Journal Proc. R. Soc. A-Math. Phys. Eng. Sci.
Volume 465 Issue 2109 Pages 2927-2948
Keywords Bayesian networks; clustering; unsupervised training; classification expectation-maximization algorithm; machine learning
Abstract This paper presents a new approach to the unsupervised training of Bayesian network classifiers. Three models have been analysed: the Chow and Liu (CL) multinets; the tree-augmented naive Bayes; and a new model called the simple Bayesian network classifier, which is more robust in its structure learning. To perform the unsupervised training of these models, the classification maximum likelihood criterion is used. The maximization of this criterion is derived for each model under the classification expectation-maximization ( EM) algorithm framework. To test the proposed unsupervised training approach, 10 well-known benchmark datasets have been used to measure their clustering performance. Also, for comparison, the results for the k-means and the EM algorithm, as well as those obtained when the three Bayesian network classifiers are trained in a supervised way, are analysed. A real-world image processing application is also presented, dealing with clustering of wood board images described by 165 attributes. Results show that the proposed learning method, in general, outperforms traditional clustering algorithms and, in the wood board image application, the CL multinets obtained a 12 per cent increase, on average, in clustering accuracy when compared with the k-means method and a 7 per cent increase, on average, when compared with the EM algorithm.
Address [Pham, Duc Truong; Ruz, Gonzalo A.] Cardiff Univ, Mfg Engn Ctr, Cardiff CF24 3AA, S Glam, Wales, Email: gonzalo.ruz@uai.cl
Corporate Author Thesis
Publisher Royal Soc Place of Publication Editor
Language English Summary Language Original Title
Series Editor Series Title Abbreviated Series Title
Series Volume Series Issue Edition
ISSN 1364-5021 ISBN Medium
Area Expedition Conference
Notes WOS:000268317700016 Approved no
Call Number UAI @ eduardo.moreno @ Serial 62
Permanent link to this record