Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

The Future of Advance Analytics


Published on

Presented at InnOTe

  • Login to see the comments

  • Be the first to like this

The Future of Advance Analytics

  1. 1. 4/18/20131The Future of Advance AnalyticsDavid SmithChief Executive Officer, HBMG Inc.dsmith@hbmginc.comRoom: Room: 102BVolume, variety, and velocity are changing us in our companies, governmentagencies and at home. How do BI, Social/Business media, mobility, Devices and BigData drive business decisions? The success is driven by the use of advanceanalytics. Business analytics facilitates realization of business objectives throughreporting of data to analyze trends, creating predictive models for forecasting andoptimizing business processes for enhanced performance. This is important not onlyin business but the military and government as well.This presentation will look at the current and future trends in analytics and how theywill impact each of us. Special emphasis will be given to the new trend of embeddedanalytics. As the world moves faster toward real-time the analytics must move aswell. Attendees will leave the session with a understanding of the future directions ofadvance analytics.The Future ofAdvanceAnalyticsAnalyticsDavid SmithCEO
  2. 2. 4/18/20132Definition for Advanced AnalyticsAnalysis is the examination process itself where analytics is the supporting technology and associated tools. BI is quite synonymous to analytics in IT context. Advanced Analytics, Business Analytics, Data Analytics, Analytics Software, Analytics Technology are almost always marketing pleonasms (redundantalways marketing pleonasms (redundant expressions) and can be safely substituted by just ‘analytics’Definition for Advanced AnalyticsAnalysis is a pretty old, well understood term d ti ll “b ki d ”and essentially means “breaking down” or “decomposition”. More accurately –the process of decomposing complex entity into simpler components for easier comprehension.
  3. 3. 4/18/20133Definition for Advanced AnalyticsAdvanced analytics provides algorithms for complex analysis of either structured or unstructured data. It includes sophisticated statistical models, machine learning, neural networks, text analytics, and other advanced data mining techniques. Among its many use cases, it can be deployed to find patterns in data, prediction, optimization, forecasting, and for complex event processing/analysis. Examples include predicting churn, identifying fraud, market basket analysis, or understanding website behavior Advanced analytics does notunderstanding website behavior. Advanced analytics does not include database query and reporting and OLAP cubes.America gets more than half its economic growth from industries that barely existed a decade ago—such is the power of innovation, especially in the information and biotechnology industries.—The EconomistCopyright, 2011 © HBMG, Inc.
  4. 4. 4/18/20134Business ProblemMore than half of business and IT executives, 56 percent, report they feel overwhelmed by the amount of data their company manages. Many report they are often delayed in making important decisions as a result of too much information. Surprisingly, 62 percent of C‐level respondents – whose time is considered the respondents whose time is considered themost valuable in most organizations – report being frequently interrupted by irrelevant incoming data. 
  5. 5. 4/18/20135Why?• Technological innovation is now the most important driver for competitive success– Many firms earn over one‐third of sales on products developed within last five years• Product life cycles ( time between product introduction to market and its withdrawal)Copyright, 2011 © HBMG, Inc.– Software 4‐12 months– Computer hardware 12‐24 months– Large home appliances 18‐36 monthsBusiness, Knowledge, and Innovation Landscape• Typically 80% of the key knowledge (and value) is held by 20% of the people – we need to get it to theheld by 20% of the people  we need to get it to the right people• Only 20% of the knowledge in an organization is typically used (the rest being undiscovered or under‐utilized)80 90% f th d t d i t d ill b• 80‐90% of the products and services today will be obsolete in 10 years – companies need to innovate & invent fasterCopyright 2012@ HBMG Inc.
  6. 6. 4/18/20136Tapping into the Data• Data Storage• Reporting• AnalyticsUtilized data• Analytics• AdvancedAnalytics– Computing withbig datasets is aUnutilized datathat can beil bl tgfundamentallydifferent challengethan doing “bigcompute” over asmall datasetavailable tobusinessInnovation:‘The real voyage of discovery consistsnot in seeing new lands,but in seeing with new eyes’(Source: Marcel Proust)‘Innovation =creative idea andimplementation’(Source: Glossary of Electronics)‘Innovation: change that‘A new method, idea,product, etc’(Source: Oxford EnglishDictionary)creates a new dimension ofperformance’(Source: Peter Drucker)‘An innovation to be effectivehas to be simple and it hasto be focused’(Source: Peter Drucker)‘Value innovators look forwhat customers value incommon’(Source: Kim & Mauborgne)‘Innovation is the process by which new products ormethods of production are introduced, including allthe steps from the inventor’s idea to bringing thenew item to market’(Source: Baumol, Economics: Principles & Policy)(Source: Peter Drucker)‘Firms need to manage steady stateinnovation and radical change becausecontinuous improvement is no longerenough’(Source: Tom Peters)
  7. 7. 4/18/20137“Big Data” and it’s close relatives “Cloud Computing”, “Social Media” and "Mobile" are the new frontier of innovation.Driven by Advance AnalyticsBig Data and It’s BrothersVolumeVolumeVarietyVelocity………..
  8. 8. 4/18/20138VolumeVolume is increasing at incredible rates. h l h h dWith more people using high speed internet connections than ever, plus these people becoming more proficient at creating content and just more people in general contributing information are combined forces thatinformation are combined forces that are causing this tremendous increase in Volume. VarietyNext in breaking down Big Data into easily digestible bite‐size chunks is the concept of Variety. Take your personal experience and think about how much information youexperience and think about how much information you create and contribute in your daily routine. Your voicemails, your e‐mails, your file shares, your TV viewing habits, your Facebook updates, your LinkedIn activity, your credit card transactions, etc. Whether you consciously think about it or not the VarietyWhether you consciously think about it or not the Variety of information you personally create on a daily basis which is being collected and analyzed is simply overwhelming. 
  9. 9. 4/18/20139VelocityThe speed at which data enters organizations these days is absolutely amazing. With mega internet bandwidth nearly being common place anymore in conjunction with the proliferation of mobile devices, this simply gives people more opportunity than ever to contribute content to storage systems. contribute content to storage systems.Worldwide digital content will double in 18 months, and every 18 months thereafter.  VELOCITYIDCCRM DataGPSDemandSpeedVelocityOpportunitiesCustomerInventoryEmailsTweetsPlanningMobileInstantMessagesVOLUME VARIETYTransactionsServiceCallsSales OrdersThingsIn 2005, humankind created 150 exabytes of information.  In 2011, over 1,200 exabytes was created.VOLUME80% of enterprise data will be unstructured, spanning traditional and non traditional sources.GartnerThe Economist
  10. 10. 4/18/201310But I Believe there are Four  V4• Volume:Gigabyte(109), Terabyte(1012), Petabyte(1015), Exabyte(1018),  Zettabytes(1021)• Variety: Structured,semi‐structured, unstructured; Text, image, audio, video, record• Velocity(Dynamic, sometimes time‐varying)• BUT needs to add and create Value!• BUT needs to add and create Value!
  11. 11. 4/18/201311Trends driving data management– The volume of data has never been greater and is growing exponentially– The value of data has never been better understood– The value of data has never been better understood– The capabilities for processing data have never been better• Higher processor performance and density are enabling advanced processing on commodity hardware• Software enhancements designed to make best use of i f d l bl hit tprocessing performance and scalable architecture• Advanced and in‐database analytics bring processing to the data, reducing latency and improving efficiency– The data deluge problem is also a big data opportunity21Copyright, 2011 © HBMG, Inc.From
  12. 12. 4/18/201312Next Generation• Cloud Computing (e.g. “Blue Cloud”)– The “network becomes the computer”D b d i ??????– Dumb devices ??????• Pervasive Computing– Monitoring and tracking almost anything– The Internet is Free• Continuous Services• The Cloud + Pervasive Computing– Smart buildingsSmart buildings– Sensor nets– “Invisible computing” or “ubiquitous computing”– Semantic Interoperability– Ad hoc networkingCopyright 2012@ HBMG Inc.Advance Analytics as a strategic asset“The future belongs to companies and people that turn data into products.”  Mike Loukides, O’Reilly24
  13. 13. 4/18/201313Advance Analytics as a strategic asset“85% of eBay’s analytic workload is new and unknown. We are architected for the unknown.”  Oliver Ratzesberger, eBay• Data exploration – data as the new oil The exploration for data, rather than the exploration of data Uncovering pockets of untapped data Processing the whole data set, without sampling eBay’s Singularity platform combines transactional data with behavioral data, enabled identification of top sellers, driving increased revenue from those sellers25Advance Analytics as a strategic asset“Groupon will not be the first or last organization to compete and win on the power of data It’s happeningcompete and win on the power of data. It s happening everywhere.”  Reid Hoffman and James SlavetGreylock PartnersData harnessing – data as renewable energyH i t ll i d t t Harnessing naturally occurring data streams Like harnessing raw energy to be converted into usable energy Conversion of raw data into usable data 26
  15. 15. 4/18/201315As the world gets smarter, infrastructure demands will growSmarttrafficsystemsSmartenergygridsSmarthealthcareSmartfoodsystemsIntelligent oilfieldtechnologiesSmart retailSmart watermanagementSmartregionsSmartweatherSmartcountriesSmartsupplychainsSmartcitiesDataDesktop
  16. 16. 4/18/201316KnowledgeCopyright 2012@ HBMG Inc.
  17. 17. 4/18/201317The Nature of Communications Has Been Changing...People toThingsPeople toPeopleThings toThingsPeople toThingsPeople toPeopleThings toThingsTOLLTOLLTOLLTOLLThingsPeople ThingsThingsPeople ThingsCopyright, 2011 © HBMG, Inc.15.62 Gallons27.33 DollarsThank You!15.62 Gallons27.33 DollarsThank You!15.62 Gallons27.33 DollarsThank You!Mainframe/TerminalClient ServerUbiquitousCOSTSMainframeMini/TerminalServer/PCASP/ApplianceUSAGEDistribution of Computing Power Virtualization UbiquitousImproved Access/ConvenienceImproved Ease of UseLarge Organizations + Small Organizations + Departments + BYOD EntrepreneursMiddle Class (PCs) + General Population
  18. 18. 4/18/201318Digital Video AdaptersSatellite RadioReceiversDigital CamerasPDAsWireless CamerasWireless TVMonitorsDigital MusicAdaptersNetworked StorageCentersGame ConsolesSmart DisplaysSmart PhonesLaptop PCsDesktop PCsWireless GamingAdaptersMovies-on-DemandReceivers“Fourth Generation”Set-top BoxesMP3 PlayersDigital Media Receivers Personal VideoRecordersNetworked DVD Player Mobile GamingDevices802.11 SpeakersCopyright, 2012 © HBMG, Inc.The Global GridCopyright 2012@ HBMG Inc.
  19. 19. 4/18/201319To Every Sensor is a ServerPhone -PDAProcessorData StorageCommunicationsRich variety ofsensorsSmart DustMicrophoneMicrostorage(Areal density 100x’s CD)Copyright, 2011 © HBMG, Inc.EmbeddedBiofluidics ChipRobot3,5004,000• MobileGrowth at the Edge of the Network1,0001,5002,0002,5003,000Petabytes/DayGlobalMobile• Device to Device• Sensors• Entertainment• Smart Home• Distributed Industrial• Autos/Trucks• Smart ToysConvergedContent05002003 2004 2005 2006 2007 2008 2009 2010 2011Year2012TraditionalComputationCopyright 2012@ HBMG Inc.
  20. 20. 4/18/201320DOD ExampleDOD Example Counterinsurgency operations are complex  increased emphasison: Information and analysis at lowest levels Shortened decision making time-scalesOperation TrendsSource: TTI Vanguard Conference - Psydexg Wider array of information sources Continued growth in volume of data, especially informal informationwith limited structure  must transform disparate info to knowledge Processing power and storage capacity increasing faster thancommunications capacity  must smartly position data and serviceswithin networks Increased use of commercial cellular networks  hybrid networks Increased use of commercial cellular networks  hybrid networksthat exploit and interoperate with commercial wireless comms is key Enhancing coalition decision making depends on securecommunications and information networks  must address end-to-end problem of data-to-decision (coalition)
  21. 21. 4/18/201321“Recce” P-38“Recce” P-38RF-101 VoodooInfo in War RevolutionTechnology—Information—OrganizationB-17 SpotterCorps0-2 Bird Dog0-2 Bird DogB-5275 Days75 Days14 Days14 Days<10 Minutes<10 MinutesN b f W N b f SRSIRIS ISR?41Evolution of Technology, Information, and Culture Enabled Move from Evolution of Technology, Information, and Culture Enabled Move from Segregation of Ops and Intel to Integration of Ops and Intel…Segregation of Ops and Intel to Integration of Ops and Intel…Number of WeaponsRequired to TargetNumber of SensorsRequired to Target19431943 20092009NationalStrategicStrategicStrategicOperationalOperationalOperationalAIRAIRSPACSPACCYBECYBEStrategicStrategicOperationalOperationalStrategicStrategicOperationalOperational21st Century Challenges:Precision and Information SynergyTacticalOperationalOperationalTacticalTacticalTacticalDESERTSTORM1991 & Prior1999ALLIEDFORCE2001ENDURINGFREEDOM2009AF/PAK & IraqTODAYINFO AGEINFO AGEWARFAREWARFARETOMORROWTOMORROWRRCECEERERTacticalTacticalppTacticalTacticalOps …+ IntelI + S&R GLOBALINTEGRATEDISRSegregatedIntel & OpsNTI+ Intel INTEGRATEDISRIntel & OpsKandahar RunwayPod RecceReal‐TimeFusionOPSINTELMulti‐DomainFusion
  22. 22. 4/18/201322More CollectorsMore CollectorsBetter SensorsBetter SensorsDimensions of ISR…“More of Everything”More DataMore Data•• More StorageMore Storage•• More CommsMore Comms•• More ToolsMore Tools•• More AnalystsMore Analysts•• More LinguistsMore LinguistsBetter Intel…All on an Operationally Responsive Timeline…All on an Operationally Responsive TimelineSensor Data VolumeHow do we handle all this data?How do we handle all this data?“Rebalancing Collection & PED may be Necessary”“Rebalancing Collection & PED may be Necessary”
  23. 23. 4/18/201323Persistent Surveillance Data RatesGeneral Norton A. SchwartzAir Force’s chief of staff “If i id f l h“If automation can provide a cue for our people that would make better use of their time, that would help us significantly,” NY Times.Lieutenant General David DeptulaFormer Deputy Chief of Staff for ISR, Headquarters, US Air p y f f ff f qForce“We’re going to find ourselves in the not too distant future swimming in sensors and drowning in data”
  24. 24. 4/18/201324Advance Analytics• Advanced Analytics and Big Data are two of the most active areas of innovation in the Tech sector• legacy infrastructures and government policies are increasingly at odds with the realities of the analytic landscape• Certain forms of analysis is no longer possible within an encrypted environment. Rules that require data to be encrypted, both while in transit and at rest, also introduce performance penalties that make it difficult if not impossible to process large datasets in an acceptable timeframeTodays CycleWhere is Real Time?
  25. 25. 4/18/201325Advance Analytics• The time to use the output is increasingly getting shorter – Real Time is becoming very commonLi it d il bl h d f i• Limited available human resources, and performance is often unreliable due to human fatigue and distraction. Therefore, automated real‐time sensor processing techniques are required to reliably detect and discriminate targets of interest• Limited automated processing and tagging tools• – Still NOT enoughAdvance Analytics• The time to use the output is increasingly getting shorter – Real Time is becoming very common• Limited available human resources, and performance is often punreliable due to human fatigue and distraction. Therefore, automated real‐time sensor processing techniques are required to reliably detect and discriminate targets of interest – Still NOT enough• Need to move to thesensor/collector• Needs to be embedded in the the sensor
  26. 26. 4/18/201326Autonomous SystemsAgents coordinateand negotiate to achievecommon goalsAgents dynamically adaptto and learn abouttheir environmentAdaptiveAutonomous Mobile InteroperateCooperativeSocialPersonalityIntelligentAgentsInformationAgentsHBMG Inc. Copyright 2012 Agents are goal directedand act on theirown performingtasks on your behalfAgents moveto where theyare neededAgents interoperatewith humans, other,legacy systems, andinformation sourcesNumbers• How many data in the world?– 800 Terabytes, 2000– 160 Exabytes, 2006– 500 Exabytes(Internet), 2009– 2.7 Zettabytes, 2012– 35 Zettabytes by 2020• How many data generated ONE day?– 7 TB, Twitter– 10 TB, FacebookBig data: The next frontier for innovation, competition, and productivityMcKinsey Global Institute 2011
  27. 27. 4/18/2013271 illi1million transactions during this presentationIf You Liked ____, You’ll Love ___ !
  28. 28. 4/18/2013281billi1billion clicks during this presentation
  29. 29. 4/18/201329Gartner Hype Cycle 2012
  30. 30. 4/18/2013302012 Business Intelligence, Analytics and Information Management Survey from InformationWeek ReportsA few insights from the report:•58% of those surveyed are “very interested” in advanced analyticsy y y•Advanced analytics is the No. 1 leading-edge technology•Cloud analytics systems are hot because they are easier on thepocketbook; yet 63% of users have privacy concerns•Data pros just can’t get good data – data quality still ranks as the topbarrier to adopting BI products throughout the company•25% of those surveyed are mobilizing their data analytics withdashboards and data visualizations40% f d t t li t t b th bi d t•40% of data pros are struggling to stay above the big data waveConclusion Data is one the major factors driving infrastructure computing The growing volume of data is a problem, but it is also an opportunity Don’t worry about ‘big data,’ worry about your datay g , y y Take a Total Data approach to data management• Think pragmatically about data storage and analysis• Attempt to capture and analyze any data that might be relevant, regardless of where it resides ‘Datastructure’ will become increasingly valuable, not only as a source of data but also as a source of intelligence The rise of the ‘data cloud’ and the PaaS data layer will encourage a more flexible approach to data management and analytics The companies that win will be those that think about data as a strategic asset and implement the technology to monetize it
  31. 31. 4/18/201331ConclusionThe World is moving to Real TimeAdvanced Analytics is the Keyy yAdvanced Analytics Must be embedded in the collectors and sensors• Think about where the data comes from• Attempt to capture and analyze any data that might be relevant, regardless of where it resides• Realize collaboration is the key in Advance Analytics just as it is in BusinessIf we don’t change ourIf we don t change ourdirection, we’ll end up exactlywhere we are headed.—Ancient Chinese Proverb
  32. 32. 4/18/201332In Parting: Be Paranoid•“Sooner or later, something fundamental in your business world•“Sooner or later, something fundamental in your business worldfundamental in your business world will change.”• Andrew S. Grove, Founder, Intel“Only the Paranoid Survive”fundamental in your business world will change.”• Andrew S. Grove, Founder, Intel“Only the Paranoid Survive”Only the Paranoid SurviveOnly the Paranoid SurviveCopyright @2008 HBMG Inc.