stay ahead of the curve Uncover new information and value to remain competitive 4 tracks designed to follow the Analysts Journey May 24 San Jose California Register at passbaconferencecom ID: 914352
Download Presentation The PPT/PDF document "Get the essential business and data anal..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Get the essential business and data analytics skills and technologies to
stay ahead of the curveUncover new information and value to remain competitive4 tracks designed to follow the Analyst’s Journey
May 2-4, San Jose, California
Register
at
passbaconference.com
today
!
(Use
discount
code
BAMARA
for
$100 off
registration)
Slide2PASS Virtual Chapters for BusinessAnalytics
www.sqlpass.org/vcFREE ONLINE LEARNING
Slide3DANIEL FYLSTRAPresident of Frontline SystemsDeveloper of Solver in Excel
25 Years in Advanced AnalyticsData/Text Mining - XLMinerOptimization – Premium SolverSimulation – Risk SolverMarketer of VisiCalc, the FirstSpreadsheet on Apple II3
www.facebook.com/FrontlineSolvers
twitter.com/
FrontlineSolver
www.linkedin.com/company/frontline-systems-inc
Slide4Advanced Analytics for Excel Users: Learn How to Do It YourselfDaniel Fylstra
, Frontline Systems
Slide55Goals for Today’s Session
Know how Analytics builds on Business IntelligenceKnow why you’d build an analytic model: business payoffsKnow what kinds of results you can get from analytic modelsKnow how you’d build your own analytic model, and howto get data into your modelKnow what to do next, if you want to learn
Slide66How Analytics Builds on Business Intelligence
“Analytics are a subset of … business intelligence: a set of technologies and processes that use data to understand business performance … The questions that analytics can answer represent the higher-value and more proactive end of this spectrum.” – Tom Davenport, Competing on Analytics
Slide77Analytics: The Three Levels
Descriptive Analytics: Classic BIQuantitative Assessment of Past Business ResultsStatistics, Exploratory Data Analysis, VisualizationPredictive AnalyticsQuantitative Methods to Predict New OutcomesForecasting, Prediction, Classification, AssociationPrescriptive AnalyticsQuantitative Methods to Make Better DecisionsDecision Trees, Monte Carlo Simulation, Optimization
Slide88Why Build Analytic Models: Example Payoffs
Two Frontline Systems Customer ExamplesExcel model to optimally deploy 83 employees with different skill sets across 24 stations saved $1.9 million per year in overtime.Excel simulation model showed major chemical company why a plant was missing goals, and how to solve the problem without any new investment.U.S. Air Force Air Logistics CenterC-5 Galaxy transport maintenance hub reduced turnaround time from 360 to 160 days,
saving taxpayers $50 million and saving soldiers’ lives.
Memorial Sloan-Kettering Cancer CenterOptimizing radiation beams reduced side-effects of treating cancer –
improving quality of life
and
saving
$
459 million per
year on prostate cancer alone.
Slide99Can This Help in Your Work or Career?
Optimization models can deliver huge cost savingsSimulation/risk analysis models can help avoid disasterBut very few business analysts have the skills to do thisIf you can do this, your value to your company will riseSome analytic models address operations, others address strategic decisionsEx. whether to build a new plant, and where to locate itBe prepared to present your work to senior management
Slide1010Descriptive Analytics: Excel & Power BI
Key task: Data access / shaping – Power Query does thisExcel + Power Pivot data model holds Past Business ResultsPivot charts, Power View, Power BI for data visualizationFormulas: Sum, Count, Average, Min, Max, Var, StdDev
Slide1111Predictive Analytics: Data Mining
Key tasks: Data shaping, applying predictive modelsData mining algorithms “fit” analytic model to past dataTrained/fitted models are applied to newly arriving dataClassify: ex. Good/Poor credit risk, Likely/Unlikely to churnPredict: ex. stock price, house price, exchange rateForecast a time series: ex. next sales from past sales historyAssociate: ex. People who bought this item also bought...Tools: Azure ML, XLMiner, Predixion, SAS, SPSS, R, others
Slide1212Prescriptive Analytics: Optimization, Simulation
Key task: Create a model – A person (you) must do thisModel must capture essential features of the business situationLarger models often get their data from BI / Descriptive AnalyticsA “What If” model is the starting point – Excel is a natural tool!Given an appropriate model, we can:Ask “What are all the possible outcomes?” – simulation/risk analysisAsk “What’s the best outcome we can achieve?” – optimizationTools: Solver,
Risk Solver, @RISK, Crystal Ball, IBM, SAS, others
Slide1313Results from an Analytic Model
Results from a data mining model:Tool to classify or predict outcomes for new casesAssessment of accuracy / predictive powerResults from a simulation model:Full range of outcomes and their likelihoodSensitivity analysis of input parameters vs. outcomesResults from an optimization model:Best attainable objective, values for decision variablesSensitivity analysis of decision variables & constraints
Slide1414Data Mining: What You Need, How You Do It
What You Need: Tools toAccess / shape data, explore / visualize dataTrain / “fit” models to data: machine learning Validate model results: statistics, Lift / ROC curves How You Do ItData “wrangling” / cleaning is usually the first stepUse feature selection to identify variables that matterTry multiple algorithms: Regression, trees, neural netsAssess and think about results: Avoid over-fitting
Slide1515Simulation: What You Need, How You Do It
What You Need: Tools toCreate a “what if” model, calculating results of interestDefine probability distributions for uncertain inputs Run Monte Carlo simulation, create statistics and charts How You Do ItDefine distributions by fitting data, or industry practiceDefine dependence among inputs: corr. matrices, copulasRun simulation, or multiple simulations with parametersAssess and think about results: stats, histograms, scatterplots
Slide1616Optimization: What You Need, How You Do It
What You Need: Tools toCreate a “what if” model, calculating results of interestDefine decision variables for inputs under your controlDefine constraints and an objective to max / minimizeRun an optimization for optimal values, sensitivity analysis How You Do ItDefine constraints for limited resources, physical conditions, policiesUnderstand dependence between outputs and inputs: linear / nonlinear Run optimization, or multiple optimizations with parameters you varyAssess and think about results: understand “dual values,” sensitivity
Slide1717Can This Help in Your Work or Career?
Optimization models can deliver huge cost savingsSimulation/risk analysis models can help avoid disasterBut very few business analysts have the skills to do thisIf you can do this, your value to your company will riseSome analytic models address operations, others address strategic decisionsEx. whether to build a new plant, and where to locate itBe prepared to present your work to senior management
Slide1818Where to Learn More: Textbooks on Amazon
Cliff Ragsdale Spreadsheet Modeling 7th EdPowell & Baker Management Science 4th EdCamm et al Essentials of Business AnalyticsJames Evans Business Analytics
Slide1919Where to Learn More: Online Courses and Tools
www.edx.orgwww.coursera.orgwww.solver.comwww.xlminer.com
Slide2020Free Tools to Get Started in Excel and Excel Online
Excel: Power Query, Power Pivot, Power View, SolverPower BI: Free account, Power BI DesignerExcel Online Office Add-ins: Solver, Risk Solver, XLMiner XLMiner.com, Rason.com: Free accounts
Slide21Thank You –See You in San Jose!Daniel Fylstra, Frontline Systems
Slide22Like What You Heard?Daniel
Fylstra will be presenting at thePASS Business Analytics Conference 2016!Pre-Conference Session (full day)Advanced Analytics for Excel Users: Learn How to Do it YourselfBreakout Session (60 min)Prescriptive Analytics: Decision Models with Real Business Payoffs
Slide2323
Access to online training and contentEnjoy discounted event ratesJoin Local Chapters and Virtual Chapters Get advance notice of member exclusives
PASS is a not-for-profit organization which offers year-round learning opportunities to data professionals
Membership is free, join today
at
www.sqlpass.org
JOIN
PASS