Extracting business value from Big Data. Dr. Rosaria Silipo Phil Winters
|
|
- Coleen Charles
- 6 years ago
- Views:
Transcription
1 Extracting business value from Big Data Dr. Rosaria Silipo Phil Winters
2 At Last years KNIME UGM Dr. Killian Thiel Dr. Tobias Kötter TEXT MINING MEETS NETWORK MINING 2
3 White Papers and complete workflows available on the KNIME Public Server! Text Mining for Sentiment Network Mining for Relevance Drill Down on special cases Analytics for Prediction 3
4 4
5 Hot Topics Telemetry Data Manipulation Time Series Analysis Combine with Predictive Analytics SENSIBLE usages of Big Data Measurable / Applied to the Business Use public data please so we can all learn 5
6 Industries with these challenges: Manufacturing Chemical Life Science Transportation Utilities Automotive Cyber Security 6
7 Energy Industry Complex Networks Regulation Green Initiatives Competition Smart Meters / Transmission 7
8 The Irish Energy Trials Smart Meters 1 Year Trial Electricity and Gas 5000 homes and businesses Full before/after Surveys 176 million Rows, 40 G of Lovely Data.. 8
9 Overview Import + Transform Data Clustering of Meter IDs Adding the Survey Results Analysing the Clusters Forecasting Electricity Usage Forecasting Peaks DataRush For KNIME Conclusions and next steps
10 Import Data 1-2
11 Clustering Meter IDs Average hourly time series cluster by cluster 30 clusters with k-means on average daily, monthly, hourly,... kw values
12 Almost Night Owls Avg daily kw Avg hourly kw size Cluster % 7% 17% 18% 20% 73 Cluster % 6% 13% 15% 13% 513 Cluster % 5% 16% 18% 25% 684 Cluster % 7% 17% 18% 24% 230 Cluster % 6% 19% 18% 20% 622 Cluster % 7% 20% 21% 19% 18 Cluster % 6% 15% 18% 29% People in these clusters use electricity more during the night than during the day 2. Cluster 10 uses really a lot of electricity all day round (Fridges? Machines?)
13 Night Owls Avg daily kw Avg hourly kw size Cluster % 7% 9% 9% 14% 58 Cluster % 7% 12% 13% 14% 59 Cluster % 7% 13% 15% 18% 30 Cluster % 7% 14% 14% 15% 439 Cluster % 6% 13% 15% 24% People in these clusters use electrcity mostly at night
14 All Rounders Avg daily kw Avg hourly kw size Cluster % 7% 21% 22% 20% 40 Cluster % 7% 15% 17% 33% 539 Cluster % 6% 20% 21% 24% 507 Cluster % 14% 18% 15% 18% 231 Cluster % 6% 25% 25% 17% 71 Cluster % 5% 18% 24% 19% 418 Cluster % 6% 21% 21% 29% 365 Cluster % 6% 29% 22% 29% 273 Cluster % 6% 29% 27% 16% These people use electricity all day round 2. Besides cluster 14 and cluster 3, they all use little electricity distributed during the day
15 Night and Day Avg daily kw Avg hourly kw size Cluster % 6% 28% 29% 19% 17 Cluster % 5% 35% 34% 15% 25 Cluster % 5% 35% 34% 15% 71 Cluster % 2% 37% 39% 14% More people use electricity at night than during the day 2. Cluster 8 uses a lot of electricity and mostly during the day 3. Cluster 9 uses very little electricity and mostly during the day
16 Survey Data available 3-4 Allocation - type if billing Residential - pre/post Survey housing type, size, age, people, insulation, attitudes, etc. SME (small and medium Enterprises) - pre /post Survey building type, size, age, industry, attitudes, etc.
17 Survey Data Survey Data Decision Tree: Time Series Cluster Predicts the Customer Type! Other Significant Fields: Industry Size in sq. m. Age of building Invested in Efficiency Owner/Renter! Meter ID and Clusters
18 Night and Day Avg daily kw Avg hourly kw size Cluster % 6% 28% 29% 19% 17 Cluster % 5% 35% 34% 15% 25 Cluster % 5% 35% 34% 15% 71 Cluster % 2% 37% 39% 14% More people use electricity at night than during the day 2. Cluster 8 uses a lot of electricity and mostly during the day 3. Cluster 9 uses very little electricity and mostly during the day
19 kw vs. hour of day cluster 8 kw Hour of day
20 Low electricity usage Cluster 8 Christmas 2009 Christmas 2010 kw Hour of day
21 On a smaller time window... Jul Weekly Periodicity Aug kw Monday Jul-20 Monday Jul-27 Tuesday Monday Aug-04 Aug-03 Bank Holiday Monday Aug-10 Monday Aug-17 Hour of day
22 On a smaller time window... Jul Weekly Periodicity Aug kw Monday Jul-20 Monday Jul-27 Monday Aug-03 Bank Holiday Hour of day
23 On a smaller time window... Wed Nov Daily Periodicity Thu Nov kw 11:00 14:00 11:00 14:00 21:00 5:00 21:00 5:00 Hour of day
24 Seasonality 12:00 12:00 16:00 5:00 Wed Jul :00 5:00 Thu Jul :00 11:00 14:00 21:00 5:00 21:00 5:00 Wed Nov Thu Nov
25 Energy and Forecasting Accuracy: What it means An improvement in forecasting accuracy of 1% was estimated to yield a saving in operating costs of approximately 10 million per year Bunn and Farmer 1984 Daily: Optimal Scheduling, Allocation Weekly: Purchase Policies, Maintenance Monthly, Yearly: Strategic Planning and Production
26 Simple Auto-Regressive Model Lag = p Transpose data from x(t) to x(t-p)... x(t-1) x(t) Correlation between x(t) and its past? Linear / Polynomial Regression of x(t) on its past MSE on test set
27 Building x(t-p),..., x(t-1), x(t) lag = p = 3 From QuickForm in Metanode
28 Auto-correlations Lag = 24 hours 24 hours Seasonality
29 Linear/Polynomial Regression Regression Target = x(t) Mean Square Error 90% training set 10% test set
30 Prediction Example Prediction from the Linear Regression:
31 Seasonality 1-24h We need a 24h template to repeat: First 24h Average 24h on training set Previous 24h Create and remove template from signal x(t+i) = x(t+i) template(i) Add template back into predictions p(t+i) = p(t+i) + template(i) Simple AR Model
32 After Seasonality Extraction Range = [-23, 7]
33 Seasonality 1 24h We need a 24h template to repeat: First 24h Average 24h on training set Previous 24h Remove template from signal x(t) = x(t) x(t-24) Add template back into predictions p(t) = p(t) + x(t-24) Simple AR Model
34 After Seasonality Extraction Range = [-19, 24]
35 Seasonality 2 24h * 7 Remove template from signal x(t) = x(t) x(t-24*7) Simple AR Model Add template back into predictions p(t) = p(t) + x(t-24*7)
36 After Seasonality Extraction Range = [-5, 7]
37 Prediction Example
38 Neural Networks Lag = p Remove template from signal x(t) = x(t) x(t-24*7) Add template back into predictions p(t) = p(t) + x(t-24*7) Data back to original range Input for NN Model must be in [0,1] NN Model MSE
39 R arima(p,d,q) R arima model including seasonality MSE
40 Regression and NN Results MSE on cluster 8 test set: MSE / Lag Seasonality Linear AR Polynomial* AR Linear AR First 24h Linear AR Previous 24h Linear AR 24h * NN 24h * *Polynomial Regression with degree = 3
41 R arima Results MSE on cluster 8 test set: MSE / (p, d, q) Seasonalit y (1, 0, 0) (2, 0, 0) (1, 0, 1) (2, 0, 1) (2, 0, 2) R arima(p,0,q) Previous 24h* *arima(gtemp, order=c(p,0,q), xreg=1:nobs, seasonal=list(order=c(p,0,q),period=24), include.mean=true) Note. The R arima -> optim procedure was failing for p > 2 or q > 2, due to memory problems.
42 R arima(1, 0, 1) Predictions
43 MSE on Peaks only MSE on cluster 8 test set only on peaks (pred(t) OR x(t) > 35 kw): MSE / Lag Seasonalit y Linear AR 24h * NN 24h *
44 Big Data
45 Import + Aggregate Data (KNIME) Several hours execution time Reading takes only 20 minutes 7h execution time: 2h datetime conversions 5h Sorter node Several hours execution time
46 Import + Aggregate Data (DataRush) 9 Minutes
47 Big Data Forecasting Clusters Posibilities: More Data or Forecasting over Meter IDs
48 Big Data Conclusions Accessing Data: Manipulating Data: Explore the Data: Predictive Mining: Execution: Real Time: Benificial Very Likely Beneficial Don t Bother Task Based! Experiment Possibly Very Likely Beneficial With KNIME and Pervasive s Rushanalyticsfor KNIME, You can mix and match as required!!!
49 Next Steps: Data Science Introduce the % weekday usage feature New clusters including % weekday usage Better R arima model Investigate more interesting clusters (cluster28 and cluster 2 of the night owls, and cluster 9 of the morning people) Introduce MA(q) model for arima(p,d,q) Implement automatic detection of p, d, and q of arima(p,d,q)
50 Next Steps: Business More Accurate Weekly, Monthly, Yearly Forecasting Pricing plans based on segments Models that relate key golden questions to predicted usage patterns We can predict the meter id s cluster! Using: Customer Type Industry Owner/Renter! Size in sq. m. Age of building Invested in Efficiency
51 Conclusion Hot Topics Telemetry Data Manipulation Time Series Analysis Combine with Predictive Analytics SENSIBLE usages of Big Data Measurable / Applied to the Business 51
52 Industries can now rock the Big Data Challenges! Manufacturing Chemical Life Science Transportation Utilities Automotive Cyber Security Slides will be available, White Paper coming! 52
Big Data, Smart Energy, and Predictive Analytics. Dr. Rosaria Silipo Phil Winters
Big Data, Smart Energy, and Predictive Analytics Dr. Rosaria Silipo Phil Winters Hot Topics Telemetry Data Time Series Analysis SENSIBLE usages of Big Data Measurable / Applied to the Business Use public
More informationR and KNIME: The Best of Two Worlds.
R and KNIME: The Best of Two Worlds. The Berkeley R Language Beginner Study Group Nov 19, 2013 Michael R. Berthold University of Konstanz, Germany KNIME.com AG, Switzerland KNIME Overview Demo / Intro
More informationAdaptive Time Series Forecasting of Energy Consumption using Optimized Cluster Analysis
Adaptive Time Series Forecasting of Energy Consumption using Optimized Cluster Analysis Peter Laurinec, Marek Lóderer, Petra Vrablecová, Mária Lucká, Viera Rozinajová, Anna Bou Ezzeddine 12.12.2016 Slovak
More informationThe 5 th KNIME User Group Meeting Welcome & Introduction. Michael R. Berthold KNIME.com AG, Zurich, Switzerland
The 5 th KNIME User Group Meeting Welcome & Introduction Michael R. Berthold KNIME.com AG, Zurich, Switzerland KNIME Meetings 1 st UGM (Konstanz): November 2007 2 nd UGM (San Francisco): February 2009
More informationCitiPower Amended Revised Proposed Tariff Structure Statement
CitiPower Amended Revised Proposed Tariff Structure Statement 2017 2020 This page is intentionally left blank. 2 CitiPower Amended Revised Proposed Tariff Structure Statement 2017 2020 Table of Contents
More informationPenn Power Load Profile Application
Penn Power Load Profile Application I. General The Company presents the raw equations utilized in process of determining customer hourly loads. These equations may be utilized by Electric Generation Suppliers
More informationPortland General Electric Company First Revision of Sheet No P.U.C. Oregon No. E-18 Canceling Original Sheet No. 26-1
First Revision of Sheet No. 26-1 P.U.C. Oregon No. E-18 Canceling Original Sheet No. 26-1 PURPOSE SCHEDULE 26 NONRESIDENTIAL DEMAND RESPONSE PILOT PROGRAM This schedule is an optional supplemental service
More informationCase Studies on Using Load Models in Network Calculation
1 / 24 D6.10.2 Case Studies on Using Load Models in Revision History Edition Date Status Editor v0.1 10.10.2012 Created A. Mutanen V1.0 13.12.2012 1 st version A. Mutanen 2 / 24 Abstract This report shows
More informationWhat s New. Bernd Wiswedel KNIME KNIME AG. All Rights Reserved.
What s New Bernd Wiswedel KNIME 2018 KNIME AG. All Rights Reserved. What this session is about Presenting (and demo ing) enhancements added in the last year By the team Questions? See us at the booth.
More informationTDWI Analytics Fundamentals. Course Outline. Module One: Concepts of Analytics
TDWI Analytics Fundamentals Module One: Concepts of Analytics Analytics Defined Data Analytics and Business Analytics o Variations of Purpose o Variations of Skills Why Analytics o Cause and Effect o Strategy
More informationPower Off & Save Pilot
Electric Ireland Customer Innovations Power Off & Save Pilot Project Progress Report 2 Ref: ENQEIR506/EI April 2018 Progress Report 2 Executive Summary The purpose of this document is to report on the
More informationPredictive Analytics
Predictive Analytics Mani Janakiram, PhD Director, Supply Chain Intelligence & Analytics, Intel Corp. Adjunct Professor of Supply Chain, ASU October 2017 "Prediction is very difficult, especially if it's
More informationResidential Real-Time Pricing: Bringing Home the Potential
Residential Real-Time Pricing: Bringing Home the Potential Kathryn Tholin Assessing the Potential for Demand Response Programs The Institute for Regulatory Policy Studies, Illinois State University May
More informationACHIEVING ENERGY CONSERVATION GOALS WITHOUT THE BROAD APPLICATION OF SMART METERS ONTARIO FEDERATION OF AGRICULTURE
ACHIEVING ENERGY CONSERVATION GOALS WITHOUT THE BROAD APPLICATION OF SMART METERS Background ONTARIO FEDERATION OF AGRICULTURE The Ontario Federation of Agriculture is a voluntary membership organization.
More informationNSPM Rate Design Pilot
NSPM Rate Design Pilot Stakeholder Meeting May 5, 2017 Agenda and Purpose Agenda Introduction of MN Pilot Development A. Liberkowski Concept and Goals Pilot Development Timeline MN Time of Use Rate Option
More informationRisk Simulation in Project Management System
Risk Simulation in Project Management System Anatoliy Antonov, Vladimir Nikolov, Yanka Yanakieva Abstract: The Risk Management System of a Project Management System should be able to simulate, to evaluate
More informationDemand Forecasting for Materials to Improve Production Capability Planning in BASF
Demand Forecasting for Materials to Improve Production Capability Planning in BASF Team 6 Raden Agoeng Bhimasta, Dana Tai, Daniel Viet-Cuong Trieu, Will Kuan National Tsing-Hua University About BASF BASF
More informationTOOL: QUESTIONNAIRE FOR ENGAGING SMES
TOOL: QUESTIONNAIRE FOR ENGAGING SMES Abstract In this tool we present a standardized questionnaire for service providers to identify potential small and medium enterprises (SMEs) for participation in
More informationKnowledgeSTUDIO. Advanced Modeling for Better Decisions. Data Preparation, Data Profiling and Exploration
KnowledgeSTUDIO Advanced Modeling for Better Decisions Companies that compete with analytics are looking for advanced analytical technologies that accelerate decision making and identify opportunities
More informationLinear model to forecast sales from past data of Rossmann drug Store
Abstract Linear model to forecast sales from past data of Rossmann drug Store Group id: G3 Recent years, the explosive growth in data results in the need to develop new tools to process data into knowledge
More informationDemand Rates - Frequently Asked Questions
Demand Rates - Frequently Asked Questions What is demand billing? Broadly speaking, demand billing is a billing rate, or tariff, under which users are charged both for the energy they use and the capacity
More informationManaging Energy Use and Cost with Integrated Metering
Managing Energy Use and Cost with Integrated Metering Regional Municipality of Durham October 29, 2015 Joe Green P.Eng, CEM Overview Duffin Creek WPCP Jointly owned by York and Durham Regions Capacity
More informationGENERAL SERVICE SCHEDULE GS-TOU OPTIONAL GENERAL SERVICE TIME OF USE METERED RATE
Delaware Electric Cooperative, Inc. Leaf No. 63 1. AVAILABILITY GENERAL SERVICE SCHEDULE GS-TOU OPTIONAL GENERAL SERVICE TIME OF USE METERED RATE Available to Members of the Cooperative for all non-residential
More informationPOTOMAC EDISON (Maryland) Load Profile Application
POTOMAC EDISON (Maryland) Load Profile Application I. General The Company presents the raw equations utilized in process of determining customer hourly loads. These equations may be utilized by Electric
More informationNew Clustering-based Forecasting Method for Disaggregated End-consumer Electricity Load Using Smart Grid Data
New Clustering-based Forecasting Method for Disaggregated End-consumer Electricity Load Using Smart Grid Data Peter Laurinec, and Mária Lucká 4..7 Slovak University of Technology in Bratislava Motivation
More information2018 Consumer Holiday Shopping Report
2018 Consumer Holiday Shopping Report With surging confidence in their own economic outlook, consumers plan a strong showing this holiday season fueled by significant shifts in how, when and where they
More informationUnderstanding Energy Efficiency Benefits from Smart Thermostats in Southern California
Understanding Energy Efficiency Benefits from Smart Thermostats in Southern California BECC Conference - Dec 2014 Ben Ho EnergyHub and Vassar College Ben Ho - BECC Dec 2014 1 30K data points per thermostat
More informationAMI Billing Regression Study Final Report. February 23, 2016
AMI Billing Regression Study Final Report February 23, 2016 Table of Contents Executive Summary... i 1 Introduction... 1 2 Random Coefficients Model... 5 2.1 RANDOM COEFFICIENTS MODEL DEVELOPMENT PROCESS...
More informationElectricity Demand László Szabó
Electricity Demand László Szabó www. erranet.org Outline What is electricity demand? Characteristics of electricity demand Technology improvement and metering Profiling Demand side management (DSM) 2 2
More informationEnergy Savings Analysis Generated by a Real Time Energy Management System for Water Distribution
Energy Savings Analysis Generated by a Real Time Energy Management System for Water Distribution Sarah Thorstensen Derceto Ltd, Auckland, New Zealand sthorstensen@derceto.com Abstract Washington Suburban
More informationJAWA Money & Markets Software
Imagine an accounting software solution that puts your business first. That serves up the smart features and fierce power you and your employees need to get the job done efficiently and brilliantly. A
More informationThe Value of Enterprise Meter Data Management Phyllis Batchelder Itron, The Netherlands
The Value of Enterprise Meter Data Management Phyllis Batchelder Itron, The Netherlands Topics The Evolution of the Value of Meter Data Business drivers for an Enterprise Meter Data Management Solution
More informationLeveraging Smart Meter Data & Expanding Services BY ELLEN FRANCONI, PH.D., BEMP, MEMBER ASHRAE; DAVID JUMP, PH.D., P.E.
ASHRAE www.ashrae.org. Used with permission from ASHRAE Journal. This article may not be copied nor distributed in either paper or digital form without ASHRAE s permission. For more information about ASHRAE,
More informationMachine Learning Models for Sales Time Series Forecasting
Article Machine Learning Models for Sales Time Series Forecasting Bohdan M. Pavlyshenko SoftServe, Inc., Ivan Franko National University of Lviv * Correspondence: bpavl@softserveinc.com, b.pavlyshenko@gmail.com
More informationSYST/OR 699 Fall 2016-Final Presentation
NOVEC Customer Segmentation Analysis Anita Ahn Mesele Aytenifsu Bryan Barfield Daniel Kim Department of Systems Engineering and Operations Research SYST/OR 699 Fall 2016-Final Presentation NOVEC Customer
More informationTime of Use Rates: A Practical Option If Done Well
Time of Use Rates: A Practical Option If Done Well November 15, 2016 Presented by Jim Lazar RAP Senior Advisor What Does This Rate Design Say? $1.50 $2.25 $2.75 2 Key Points TOU rates are appropriate if:
More informationPower Options. For Oregon Customers. Choosing an Electricity Service Supplier About transition adjustments... 7
P A C I F I C P O W E R Power Options 2011 For Oregon Customers Power Options At a Glance... 1 Direct Access... 3 10 Getting started... 4 Important dates... 4 The enrollment process... 5 Choosing an Electricity
More informationA Prediction Reference Model for Air Conditioning Systems in Commercial Buildings
A Prediction Reference Model for Air Conditioning Systems in Commercial Buildings Mahdis Mahdieh, Milad Mohammadi, Pooya Ehsani School of Electrical Engineering, Stanford University Abstract Nearly 45%
More informationConsumer flash estimates
European Commission Directorate General Economic and Financial Affairs Consumer flash estimates Roberta Friz EU WORKSHOP ON RECENT DEVELOPMENTS IN BUSINESS AND CONSUMER SURVEYS Brussels, 14 October 2009
More informationApproaching an Analytical Project. Tuba Islam, Analytics CoE, SAS UK
Approaching an Analytical Project Tuba Islam, Analytics CoE, SAS UK Approaching an Analytical Project Starting with questions.. What is the problem you would like to solve? Why do you need analytics? Which
More informationProblem 5: Forecasting the demand for bread. V UCM Modelling Week Master in Mathematical Engineering UCM
V UCM Modelling Week Master in Mathematical Engineering UCM 0. Summary 1. Introduction 2. Descriptive analysis 3. Classification 4. Data imputation for missing and censured values 5. Preliminary model
More informationARCHITECTURES ADVANCED ANALYTICS & IOT. Presented by: Orion Gebremedhin. Marc Lobree. Director of Technology, Data & Analytics
ADVANCED ANALYTICS & IOT ARCHITECTURES Presented by: Orion Gebremedhin Director of Technology, Data & Analytics Marc Lobree National Architect, Advanced Analytics EDW THE RIGHT TOOL FOR THE RIGHT WORKLOAD
More informationInformatics solutions for decision support regarding electricity consumption optimizing within smart grids
BUCHAREST UNIVERSITY OF ECONOMIC STUDIES Doctoral School of Economic Informatics Informatics solutions for decision support regarding electricity consumption optimizing within smart grids SUMMARY OF DOCTORAL
More informationinteliscaler Workload and Resource Aware, Proactive Auto Scaler for PaaS Cloud
inteliscaler Workload and Resource Aware, Proactive Auto Scaler for PaaS Cloud Paper #10368 RS Shariffdeen, UKJU Bandara, DTSP Munasinghe, HS Bhathiya, and HMN Dilum Bandara Dept. of Computer Science &
More informationNovel model for defining electricity tariffs using residential load profile characterisation
Int. J. xxxxxxxxx xxxxxxxxxxxxxms, Vol. X, No. Y, xxxx 1 Novel model for defining electricity tariffs using residential load profile characterisation Sima Davarzani 1, Ioana Pisica 1 and Laurentiu Lipan
More informationISO New England 2003 Demand Response Programs
ISO New England 2003 Demand Response Programs Presented at: National Accounts Demand Response Seminar August 19, 2003 Bob Laurita Sr. Program Administrator, Demand Response ISO New England, Inc. New England
More informationWhat lies below the curve
What lies below the curve Claude Godin, Principal Strategist, Policy Advisory and Research 1 SAFER, SMARTER, GREENER About DNV GL Founded 1864 Headquartered in Norway 10,000 employees Strong position in:
More informationData. Does it Matter?
Data. Does it Matter? Jarut N. Cisco Systems Data & Analytics are Top of Mind in Every Industry Automotive Auto sensors reporting location, problems Communications Location-based advertising Consumer
More informationCapacity Performance Training. June 24, 2015
Capacity Performance Training June 24, 2015 Training Objectives Provide Capacity Market Sellers with information necessary to participate in the Reliability Pricing Model (RPM) under a Capacity Performance
More informationBetter Understanding Customers: Developing SMB DNA to Improve Customer Interactions and Catalyze Positive Behavior Changes
Better Understanding Customers: Developing SMB DNA to Improve Customer Interactions and Catalyze Positive Behavior Changes Anne-Lise Laurain, Tingwen Bao, Pawel Zawadzki, Kevin Johnson, Pacific Gas & Electric
More informationWorking Together to Manage Your Company s Energy Use.
Energy for What s Ahead SM Working Together to Manage Your Company s Energy Use. Take advantage of one or more of our Demand Response (DR) programs to help lower your energy costs when you actively reduce
More informationPreliminary Analysis of High Resolution Domestic Load Data
Preliminary Analysis of High Resolution Domestic Load Data Zhang Ning Daniel S Kirschen December 2010 Equation Chapter 1 Section 1 1 Executive summary This project is a part of the Supergen Flexnet project
More informationIntroduction of Energy Data Analysis Center(EDAC)
Introduction of Energy Data Analysis Center(EDAC) Kyung-Soon Park 2017. 6. 7. Table of contents Building Energy Data & BEMS BEMS Dissemination Scheme(KEA s EDAC) Current Status and Future Plans 1 Building
More informationFinding Hidden Intelligence with Predictive Analysis of Data Mining
Finding Hidden Intelligence with Predictive Analysis of Data Mining Rafal Lukawiecki Strategic Consultant, Project Botticelli Ltd rafal@projectbotticelli.com Objectives Show use of Microsoft SQL Server
More informationLoad Management with Ripple Control Technology
Load Management with Ripple Control Technology 1. Introduction Reducing cost is a constant challenge for all energy supplying companies. A significant part of the expenditures for energy are performance-related
More information2017 Fall Summit Opening and Introduction
2017 Fall Summit Opening and Introduction Michael Berthold KNIME #KNIMESummit2017 2017 KNIME AG. All Rights Reserved. Noteworthy News 2017 KNIME AG. All Rights Reserved. 2 2017 News March: 20M investment
More informationFINANCIAL MANAGEMENT FREE
FINANCIAL MANAGEMENT FREE Greentree FREE - Financial Reporting Engine in Excel Accountants love Microsoft Excel, and many use this tool every day to format, plan, and manipulate information for a wide
More informationQuick Start Guide (for PacifiCorp Customers) January, 2011
Quick Start Guide (for PacifiCorp Customers) January, 2011 Contents Chapter 1 Signing On to Energy Profiler Online 2 Chapter 2 Overview of Analysis Capabilities.. 3 Chapter 3 Selecting Accounts/Groups.....
More informationSmart Metering and the Need for Advanced Data Management
Smart Metering and the Need for Advanced Data Management Brian Owenson Sr Director, Technology Strategy, Oracle Utilities Global Business Unit 1 Agenda A Little History Some Definitions Current State of
More informationSeasonal ladjustment of Economic Time Series
Seasonal ladjustment of Economic Time Series Revenue Estimating Training Workshop Retreat October 30, 2012 Office of Economic & Demographic Research 1 What Is a Time Series? A time series is a sequence
More informationVTWAC Project: Demand Forecasting
: Demand Forecasting IEEE PES Green Mountain Chapter Rutland, Vermont, 23 June 2016 Mathieu Sinn, IBM Research Ireland 1 Outline Smarter Energy Research in IBM Background & demo Data sources Analytics
More informationForecasting the revenue generated by ATM / CARD / CASH for EZTABLE to identify the potential / actual revenue from different payment systems.
Forecasting the revenue generated by ATM / CARD / CASH for EZTABLE to identify the potential / actual revenue from different payment systems. Team #6 Member Sean, Xie 103078517 Sam, Wang 103078502 Lydia,
More informationCutting Peak Demand Two Competing Paths and Their Effectiveness
Cutting Peak Demand Two Competing Paths and Their Effectiveness Tony Larson, National Grid, Waltham, MA Mona Chandra, National Grid, Waltham, MA Kathleen Ward, Navigant, Boulder, CO Debbie Brannan, Navigant,
More informationKnowing is half the battle: How emerging energy information systems can drive and verify savings
Data Driven Approaches To Optimizing Building Energy Performance Knowing is half the battle: How emerging energy information systems can drive and verify savings Jessica Granderson Lawrence Berkeley National
More informationPeaking Interest: How awareness drives the effectiveness of time-of-use electricity pricing
Peaking Interest: How awareness drives the effectiveness of time-of-use electricity pricing Brian C. Prest Ph.D. Candidate, Duke University brian.prest@duke.edu USAEE 2017 Annual Conference November 13,
More informationCluster-based Forecasting for Laboratory samples
Cluster-based Forecasting for Laboratory samples Research paper Business Analytics Manoj Ashvin Jayaraj Vrije Universiteit Amsterdam Faculty of Science Business Analytics De Boelelaan 1081a 1081 HV Amsterdam
More informationDynamic Pricing Works in a Hot, Humid Climate
Dynamic Pricing Works in a Hot, Humid Climate Evidence from Florida BY AHMAD FARUQUI, NEIL LESSEM, SANEM SERGICI 30 PUBLIC UTILITIES FORTNIGHTLY MAY 2017 XT here is some debate about the efficacy of dynamic
More informationIBM SPSS Statistics: What s New
: What s New New and enhanced features to accelerate, optimize and simplify data analysis Highlights Extend analytics capabilities to a broader set of users with a cost-effective, pay-as-you-go software
More informationSales Forecast for Rossmann Stores SUBMITTED BY: GROUP A-8
Sales Forecast for Rossmann Stores SUBMITTED BY: GROUP A-8 Executive Summary: a. Problem description: Business Problem: Rossman is Germany s second largest drug store chain with more than 1000 stores across
More informationCity of Ames Electric. City of Ames Electric. Smart Energy. Services. Donald Kom Director
City of Ames Electric City of Ames Electric Services Smart Energy Donald Kom Director City of Ames City of Ames Electric Services City of Ames City of Ames Electric Services Electric Services has the ability
More informationChristian Johansson, Global Product Manager Decathlon Software ABB Decathlon Software. AS Systemintegratörer
Christian Johansson, Global Product Manager Decathlon Software ABB Decathlon Software AS2016 - Systemintegratörer April 14, 2016 Plant operations Make the right decisions in a competitive environment April
More informationWest Penn Power Load Profile Application
West Penn Power Load Profile Application I. General The Company presents the raw equations utilized in process of determining customer hourly loads. These equations may be utilized by Electric Generation
More informationForecasting Seasonal Footwear Demand Using Machine Learning. By Majd Kharfan & Vicky Chan, SCM 2018 Advisor: Tugba Efendigil
Forecasting Seasonal Footwear Demand Using Machine Learning By Majd Kharfan & Vicky Chan, SCM 2018 Advisor: Tugba Efendigil 1 Agenda Ø Ø Ø Ø Ø Ø Ø The State Of Fashion Industry Research Objectives AI In
More informationForecasting fruit demand: Intelligent Procurement
2012 Forecasting fruit demand: Intelligent Procurement FCAS Final Project Report Predict fruit sales for a 2 day horizon to efficiently manage procurement logistics. Dinesh Ganti(61310071) Rachna Lalwani(61310845),
More informationSAP Predictive Analytics Suite
SAP Predictive Analytics Suite Tania Pérez Asensio Where is the Evolution of Business Analytics Heading? Organizations Are Maturing Their Approaches to Solving Business Problems Reactive Wait until a problem
More informationConsumer Segmentation and Knowledge Extraction from Smart Meter and Survey Data
Consumer Segmentation and Knowledge Extraction from Smart Meter and Survey Data Tri Kurniawan Wijaya 1*, Tanuja Ganu 2, Dipanjan Chakraborty 2, Karl Aberer 1, Deva P. Seetharam 2 1) EPFL, Switzerland &
More informationGSAW 2018 Machine Learning
GSAW 2018 Machine Learning Space Ground System Working Group Move the Algorithms; Not the Data! Dan Brennan Sr. Director Mission Solutions daniel.p.brennan@oracle.com Feb, 2018 Copyright 2018, Oracle and/or
More informationPOWER OPTIONS. For Oregon Customers. Power Options At a Glance 1. Direct Access Getting started...4. Important dates...4
P A C I F I C P O W E R POWER OPTIONS For Oregon Customers Power Options At a Glance 1 Direct Access...3 10 Getting started...4 Important dates...4 The enrollment process...5 Choosing an Electricity Service
More informationForecasting models for short and long term gas price
Forecasting models for short and long term gas price A Data Science point of view 2nd AIEE Energy Symposium, Rome, Italy November 3, 2017 Introduction Macroeconomic, Energy Analysis and Forecasting (MEAF)
More informationELECTRIC LOAD FORECAST 2010/11 to 2030/31
ELECTRIC LOAD FORECAST 2010/11 to 2030/31 (For External Use Only) IMPORTANT: THIS MATERIAL IS THE EXCLUSIVE PROPERTY OF MANITOBA HYDRO AND ALL RIGHTS ARE RESERVED. ANY RELEASE, REPRODUCTION OR OTHER USE
More informationENERGY MARKET UPDATE October 9, 2014
ENERGY MARKET UPDATE October 9, 2014 Winter is Right Around the Corner Are You Ready? Six months ago one of the most brutal winters on record in the US was winding down. Many energy consumers, who had
More informationCorrelation and Instance Based Feature Selection for Electricity Load Forecasting
Correlation and Instance Based Feature Selection for Electricity Load Forecasting Irena Koprinska a, Mashud Rana a, Vassilios G. Agelidis b a School of Information Technologies, University of Sydney, Sydney,
More informationCiti Bike. Modeling the Relationship between Earned Media Activity and Service Engagement. Allyson Hugley TAMU Analytics 2017 March 2017
Citi Bike Modeling the Relationship between Earned Media Activity and Service Engagement Allyson Hugley TAMU Analytics 2017 March 2017 1 Table of Contents Executive Summary pp. 3-7 Data & Data Sources
More informationMarket Settlements - Advanced
Market Settlements - Advanced Transmission Services Module PJM State & Member Training Dept. PJM 2017 Agenda Transmission Services Billing Examples Point-To-Point PJM 2017 2 Transmission Service Network
More informationEnergy Performance Indicators and Benchmarking Neil Brown, Paul Fleming
Energy Performance Indicators and Benchmarking Neil Brown, Paul Fleming Institute of Energy and Sustainable Development About IESD Definitions Contents Energy Saving Measures Benchmarking in detail Some
More informationForecasting With History
Forecasting With History Santiago Gallino Tuck School of Business Toni Moreno Kellogg School of Management January 2017 July 2013 LBS London, UK Learning Modules 1. Demand forecasting 2. Inventory Decisions
More informationIntroducing Analytics with SAS Enterprise Miner. Matthew Stainer Business Analytics Consultant SAS Analytics & Innovation practice
Introducing Analytics with SAS Enterprise Miner Matthew Stainer Business Analytics Consultant SAS Analytics & Innovation practice FROM DATA TO DECISIONS Optimise Competitive Advantage What is the best
More informationFORECASTING PV MARKET ADOPTION AND DEMAND IMPACTS
14TH ANNUAL ENERGY FORECASTING MEETING / EFG SCOTTSDALE, ARIZONA MAY 18-20, 2016 FORECASTING PV MARKET ADOPTION AND DEMAND IMPACTS MIKE RUSSO INSTALLED SOLAR CAPACITY New Jersey Installed capacity of 1,000
More informationSt Louis CMG Boris Zibitsker, PhD
ENTERPRISE PERFORMANCE ASSURANCE BASED ON BIG DATA ANALYTICS St Louis CMG Boris Zibitsker, PhD www.beznext.com bzibitsker@beznext.com Abstract Today s fast-paced businesses have to make business decisions
More informationA FRAMEWORK FOR CAPACITY ANALYSIS D E B B I E S H E E T Z P R I N C I P A L C O N S U L T A N T M B I S O L U T I O N S
A FRAMEWORK FOR CAPACITY ANALYSIS D E B B I E S H E E T Z P R I N C I P A L C O N S U L T A N T M B I S O L U T I O N S Presented at St. Louis CMG Regional Conference, 4 October 2016 (c) MBI Solutions
More informationPECO Backcasting Methodology Effective 2/13/2015
PECO Backcasting Methodology Effective 2/13/2015 Version 2.0, Revised 2/9/2015 The backcasting process is based on the PECO load shapes available on the SUCCESS web site under general reports Backcasting/Load
More informationSAS Visual Statistics 8.1: The New Self-Service Easy Analytics Experience Xiangxiang Meng, Cheryl LeSaint, Don Chapman, SAS Institute Inc.
ABSTRACT Paper SAS5780-2016 SAS Visual Statistics 8.1: The New Self-Service Easy Analytics Experience Xiangxiang Meng, Cheryl LeSaint, Don Chapman, SAS Institute Inc. In today's Business Intelligence world,
More informationAutomated Demand Response using the Internet of Things An Opportunities Assessment
Automated Demand Response using the Internet of Things An Opportunities Assessment Ramesh Hariharan CompuSharp Inc. Aug 17, 2016 Today: we will discuss Power Systems Operation and Control Automatic Demand
More informationPreface to the third edition Preface to the first edition Acknowledgments
Contents Foreword Preface to the third edition Preface to the first edition Acknowledgments Part I PRELIMINARIES XXI XXIII XXVII XXIX CHAPTER 1 Introduction 3 1.1 What Is Business Analytics?................
More informationSAS Machine Learning and other Analytics: Trends and Roadmap. Sascha Schubert Sberbank 8 Sep 2017
SAS Machine Learning and other Analytics: Trends and Roadmap Sascha Schubert Sberbank 8 Sep 2017 How Big Analytics will Change Organizations Optimization and Innovation Optimizing existing processes Customer
More informationCOST REFLECTIVE TARIFF (CRT) Rates as per (2018)
COST REFLECTIVE TARIFF (CRT) Rates as per (2018) INTRODUCTION On Thursday 13 October 2016, it was announced by the AER that the Council of Ministers had approved the introduction of a new tariff for high
More informationGetting the Most out of Statistical Forecasting!
Getting the Most out of Statistical Forecasting! Author: Ryan Rickard, Senior Consultant Published: July 2017 About SCMO 2 Founded in 2001, SCMO2 Specializes in High-End Supply Chain Consulting Work Focused
More informationDemand Response Programs. Donnie Clary CoServ President/CEO
Demand Response Programs Donnie Clary CoServ President/CEO CoServ at a Glance Counties served: Collin, Cooke, Denton, Grayson, Tarrant and Wise Total meters: Electric 215,000 Gas - 108,000 Electric load:
More informationElectric Forward Market Report
Mar-01 Mar-02 Jun-02 Sep-02 Dec-02 Mar-03 Jun-03 Sep-03 Dec-03 Mar-04 Jun-04 Sep-04 Dec-04 Mar-05 May-05 Aug-05 Nov-05 Feb-06 Jun-06 Sep-06 Dec-06 Mar-07 Jun-07 Sep-07 Dec-07 Apr-08 Jun-08 Sep-08 Dec-08
More informationInfor CPM d/epm (Roadmap)
Infor CPM d/epm (Roadmap) Dhruv Parekh, Solution Consultant, Infor 1 Copyright 2014 Infor. All rights reserved. Disclaimer This document reflects the direction Infor may take with regard to the specific
More information