Seite - 137 - in Intelligent Environments 2019 - Workshop Proceedings of the 15th International Conference on Intelligent Environments

themodel predicts that there are no low temperatures and really there are), are 1%and areconsideredamodelquiteoptimistic tobe implanted. As futurework, newvariableswill be incorporated into theLSTMtocreate amul- tivariateLSTMand study the influenceof other variables on temperature prediction, as wellas thenetworkadjustmentcreatinganewarchitecturewithmorelayersanddifferent activation functionsanddifferent learning factors. Acknowledgment Thisworkwas partially supported by theFundacinSe´neca delCentro deCoordinacio´n de la Investigacio´n de la Regio´n de Murcia under Project 20813/PI/18, and by Span- ish Ministry of Science, Innovation and Universities under grants TIN2016-78799-P (AEI/FEDER,UE)andRTC-2017-6389-5. References [1] Khalid A Al-Gaadi, Abdalhaleem A Hassaballa, ElKamil Tola, Ahmed G Kayad, Rangaswamy Madugundu,BanderAlblewi, andFahadAssiri. Predictionofpotatocropyieldusingprecisionagricul- ture techniques.PloSone, 11(9):e0162219,2016. [2] KMBarlow,BPChristy,GJOleary,PARiffkin, and JGNuttall. Simulating the impact of extremeheat and frost eventsonwheat cropproduction:Areview.FieldCropsResearch, 171:109–119,2015. [3] DavidBretreger, JuanQuijano, JohnAwad, et al. Monitoring irrigationvolumesusingclimatedataand remote sensingobservations. InHydrologyandWaterResourcesSymposium(HWRS2018):Waterand Communities, page112.EngineersAustralia, 2018. [4] IanGoodfellow,YoshuaBengio, andAaronCourville.Deep learning. MITpress, 2016. [5] MA´ngelGuille´n-Navarro,RaquelMartı´nez-Espan˜a,Bele´nLo´pez, and Jose´M.Cecilia. Ahighperfor- mance iot solution to reducefrostdamages instonefruits.ConcurrencyandComputation:Practiceand Experience, 2019. [6] MA´ngelGuille´n-Navarro,FernandoPeren˜ı´guez-Garcı´a,andRaquelMartı´nez-Espan˜a. Iot-basedsystem to forecast crop frost. In2017 InternationalConferenceon IntelligentEnvironments (IE), pages28–35. IEEE,2017. [7] AhmadNizarHarun,MohamedRawideanMohdKassim, IbrahimMat, and Siti SarahRamli. Preci- sion irrigationusingwireless sensor network. In2015 InternationalConferenceonSmart Sensors and Application (ICSSA), pages71–75. IEEE,2015. [8] SeppHochreiter and Ju¨rgenSchmidhuber. Long short-termmemory. Neural computation, 9(8):1735– 1780,1997. [9] SeppHochreiter andJu¨rgenSchmidhuber. Lstmcansolvehard long time lagproblems. InAdvances in neural informationprocessingsystems, pages473–479,1997. [10] AndreasKamilarisandFrancescXPrenafeta-Boldu.Deeplearning inagriculture:Asurvey.Computers andElectronics inAgriculture, 147:70–90,2018. [11] Frederik Kratzert, Daniel Klotz, Claire Brenner, Karsten Schulz, and Mathew Herrnegger. Rainfall- runoff modelling using long-short-term-memory (lstm) networks. Hydrol. Earth Syst. Sci. Discuss., https://doi. org/10.5194/hess-2018-247, in review, 2018. [12] Anne-KatrinMahlein. Plant disease detection by imaging sensors–parallels and specific demands for precisionagricultureandplantphenotyping.PlantDisease, 100(2):241–251,2016. [13] XanthoulaEirini Pantazi,DimitriosMoshou,ThomasAlexandridis,RLWhetton, andAbdulMounem Mouazen.Wheatyieldpredictionusingmachine learningandadvancedsensing techniques.Computers andElectronics inAgriculture, 121:57–65,2016. [14] Emanuele Pierpaoli,GiacomoCarli, Erika Pignatti, andMaurizioCanavari. Drivers of precision agri- culture technologiesadoption: a literature review.ProcediaTechnology, 8:61–69,2013. [15] JagadeeshDPujari,RajeshYakkundimath, andAbdulmunafSByadgi. Imageprocessingbaseddetec- tionof fungaldiseases inplants.ProcediaComputerScience, 46:1802–1808,2015. M.Á.Guillén-Navarroetal. /AnLSTMDeepLearningScheme forPredictionofLowTemperatures 137