WIXLIB

Predictive Analytics and MachineLearning for UtilitiesАндрій ТищенкоSAPPUBLIC Ukraine

DisclaimerThe information in this presentation is confidential and proprietary to SAP and may not be disclosed without the permission of SAP.Except for your obligation to protect confidential information, this presentation is not subject to your license agreement or any other serviceor subscription agreement with SAP. SAP has no obligation to pursue any course of business outlined in this presentation or any relateddocument, or to develop or release any functionality mentioned therein.This presentation, or any related document and SAP's strategy and possible future developments, products and or platforms directions andfunctionality are all subject to change and may be changed by SAP at any time for any reason without notice. The information in thispresentation is not a commitment, promise or legal obligation to deliver any material, code or functionality. This presentation is providedwithout a warranty of any kind, either express or implied, including but not limited to, the implied warranties of merchantability, fitness for aparticular purpose, or non-infringement. This presentation is for informational purposes and may not be incorporated into a contract. SAPassumes no responsibility for errors or omissions in this presentation, except if such damages were caused by SAP’s intentional or grossnegligence.All forward-looking statements are subject to various risks and uncertainties that could cause actual results to differ materially fromexpectations. Readers are cautioned not to place undue reliance on these forward-looking statements, which speak only as of their dates,and they should not be relied upon in making purchasing decisions. 2017 SAP SE or an SAP affiliate company. Allrights reserved. ǀPUBLIC2

Predictive Analytics and Machine LearningAnalytics / BIAggregateDataVisualizeHumans make decisionData is aggregated forvisualizationUI integration at bestTrainModelPrepareDataApplyModelMachines propose/makedecisionData is de-normalized,flattened, fine grainProcess integration at lastMonitorAdvanced Analytics / Predictive Analytics 2017 SAP SE or an SAP affiliate company. Allrights reserved. ǀPUBLIC3

Data Science & Machine Learning PortfolioData Scientists & Citizen Data ScientistsLine of Business UserSAP Analytics CloudData ManagerAutomated ModelerExpert Modeler(Visual Composition Framework)Predictive FactorySAP Predictive AnalyticsSpatialDB2OracleTeradata etc StreamingAnalyticsSAP ApplicationsGraphSeriesDataPredictive(PAL/APL)HANA Predictive & Machine LearningHadoop / SparkVoraDBTextAnalyticsOnPremiseSAP HANAHadoop 2017 SAP SE or an SAP affiliate company. Allrights reserved. ǀPUBLICSAP Fraud ManagementDevelopers andData ScientistsFunctionalservicesBusiness servicesHANA DBaaSLeonardo Machine Learning FoundationBig Data ServicesSAP Cloud Platform4

SAP Predictive AnalyticsCore workflowsPrepare Datawith Data Manager Build analytical datasets with clicks, not code Create thousands of derived features to increasepredictive accuracyBuild robust Predictive Models quicklywith Automated Modeler Automate dataset production & create reusabletransformations Identify which variables are changing over time withtimestamped populations Generate automatically reusable SQL code withassociated documentation 2017 SAP SE or an SAP affiliate company. Allrights reserved. ǀPUBLICBuild complex Predictive Pipelineswith Expert Analytics Easy to Use - Drag-and-drop data selection,preparation and predictive modelling Use the predictive models in SAP HANA such asUnified Demand Forecast (UDF), PredictiveAnalytics Library (PAL) & APLAutomate executive and operational reports In Database Execution – Automated PredictiveLibrary (APL) on SAP HANA & Native SparkModelling on HadoopLeverage 8000 existing R functions andlibraries Embed the models in external SAP applications Automate Predictive Modelling withClassification, Regression, Clustering, TimeSeries Forecast, Association Rules Identify automatically of key contributingvariables on very wide datasets 5

SAP Predictive AnalyticsCore workflowsLink Analysis &Recommendation Extract variables for enhanced link analysisand prediction Identify communities amongst yourcustomers Find influencers within communities to focusefforts where they count the most Create personalized recommendations foreach visitor 2017 SAP SE or an SAP affiliate company. Allrights reserved. ǀPUBLICScoringOperationalization withPredictive Factory In-database scoring using SQL Interface with business applications usingscoring equations and code: SQL, Java,PMML, SAS, C, C Manage lifecycle of thousands of models inparallel, whatever their origin (AutomatedModeler & Expert Analytics) Real Time Scoring on SAP HANA andSpark Streaming environmentsSchedule model automated application to newdata Detect data deviation & retrain modelautomatically when required Event and time based scheduling Segmented Time Series Modelling 6

Two complementary approaches to Machine LearningAnswering Experts and Business AnalystsComplementaryModeling ApproachsAutomated Modeler Guided Workflow for all users – No coding (low touch userexperience) Maximize user productivity through totally automated modelselection Support all major Data Mining functions (Time Series, Classification,Regression, Clustering, Link Analysis and Recommendation) In-database apply on all platforms In-database training on SAP HANA & Hadoop All data sources for training & apply (ODBC connections) 2017 SAP SE or an SAP affiliate company. Allrights reserved. ǀPUBLICExpert Analytics Predictive Pipelines for advanced users (high touch user experience) Build Predictive Pipelines using a drag and drop editor with support for nativepredictive libraries including PAL, APL & R on SAP HANA Advanced configuration of Automated Modeler Nodes (APL) Ready for Massive Modelling including support for Data Manager (AutomatedData Preparation) & Predictive Factory (Industrialization) Self-sufficient end to end user experience for Data Scientists with support forDatasets prepared in Data Manager7

Machine Learning in ApplicationsAccurate predictions and advanced analytics insights for business decision makersSAP Predictive AnalyticsPredictive Analytics Integrator (PAI)SAP S/4HANAPublish ModelsSAP Analytics CloudC4CPrepare & refine modelsMachine Learning platform for Data Scientists andBusiness Analysts to create, manage and monitorpredictive models to answer to business issues 2017 SAP SE or an SAP affiliate company. Allrights reserved. ǀPUBLICFramework for SAP Developersto embed Machine Learning algorithmsin SAP Applications8

SAP Predictive Analytics Integrator (PAI)Machine Learning Framework forSAP Applications PAI enables applications to deliver preconfiguredpredictive use cases to customers. Scores are consumed directly in the LOB application bythe decision maker. The lifecycle of the models within the application iseasily managed, include model retraining. Lengthy development cycles are avoided. Customers can seamlessly embedded moresophisticated models or create their own predictive usecases using SAP Predictive Analytics. 2017 SAP SE or an SAP affiliate company. Allrights reserved. ǀPUBLIC9

Predictive Analytics use cases for Utilities Моделирование и прогнозирование загрузки сети/потребления Прогнозирование отключений/перебоев/потерь Оптимизация технического и регламентного обслуживания Энерго-эффективность и аудит Анализ показаний «Смарт счетчиков» и выявление несоответствий Оптимизация резервирования Построение моделей для объяснения и прогноза условно постоянных иусловно переменных потерь активной мощности для энергосистемы Моделирование зависимости потерь от: метеоданных, ремонтов,транзитных перетоков по сечениям, напряжения, напряжения вконтрольных точках PUBLICПрогнозирование цен на ЭЭ в рамках новой модели рынка

Проект в Энергосбытовой компании 2017 SAP SE or an SAP affiliate company. Allrights reserved. ǀPUBLIC11

Результаты применения SAP Predictive Analytics вОАО Нефтедобывающая Компания 2017 SAP SE or an SAP affiliate company. Allrights reserved. ǀPUBLIC12

Прогнозирование иминимизация потерьПроект в Федеральной Сетевой КомпанииPUBLIC

Бизнес-кейс: Прогнозирование объема потребленияэлектроэнергии клиентами ое (на 1-2 дня вперед) прогнозирование общего объема потребления э/эклиентами энергосбытовой компании (ЭСК) необходимо ЭСК для формирования заявокна покупку э/э на РСВ.РеализацияДемо-пример был реализован на реальных данных ЭСК за 2010 год. В качествеисходных данных использовались исторические данные за 2010 год о почасовомпотреблении э/э клиентами ЭСК, а также почасовые данные о погодных условиях врегионе обслуживания ЭСК. Необходимо было построить прогнозною модель, котораябы строила прогноз потребления э/э клиентами ЭСК на следующие 2 дня на основанииисторических данных о потреблении э/э и на основе прогноза погоды на эти 2 дня.РезультатыПрогнозная модель, построенная SAP PA, показала высокую точность прогнозирования.Средняя ошибка прогнозирования составила 1,41% при средней допустимой ошибке 3%. 2017 SAP SE or an SAP affiliate company. Allrights reserved. ǀPUBLIC14

Бизнес-кейс: исторические данные (обучающаявыборка) 2017 SAP SE or an SAP affiliate company. Allrights reserved. ǀPUBLIC15

Бизнес-кейс: фактический объем потребленияэлектроэнергии клиентами ЭСК за январь – август2010 года 2017 SAP SE or an SAP affiliate company. Allrights reserved. ǀPUBLIC16

Бизнес-кейс: результат прогноза – сравнениефактического и предсказанного объемов потребленияэ/э 300Фактический объем э/э 2017 SAP SE or an SAP affiliate company. Allrights reserved. ǀPUBLICПредсказанный объем э/э17