If I can't picture it, I can't understand it. Albert Einstein ADVANCED DATA VISUALIZATION REDUCE TO THE TIME TO INSIGHT AND DRIVE DATA DRIVEN DECISION MAKING Mark Wolff, Ph.D. Principal Industry Consultant Health and Life Sciences Global Practice SAS Institute HISTORY JOHN W. TUKEY The greatest value of a picture is when it forces us to notice what we never expected to see. John Tukey Automatic analysis techniques such as statistics and data mining developed independently from visualization and interaction techniques Important shift from confirmatory data analysis (using exploratory data analysis charts and other visual representations to just present results) to exploratory data analysis (interacting with the data/results) J.W. Tukey. Exploratory Data Analysis. Addison-Wesley, Reading MA, 1977. 1
DEFINITIONS VISUAL The science of analytical reasoning facilitated by interactive human-machine interfaces" Visual analytics combines automated analysis with interactive visualizations for an effective understanding, reasoning and decision making on the basis of very large and complex datasets. P. C. Wong and J. Thomas. Visual analytics. IEEE Computer Graphics and Applications, 24(5):20 21, 2004. Daniel Keim, Jörn Kohlhammer, Geoffrey Ellis and Florian Mansmann (Eds.) Mastering The Information Age Solving Problems with Visual Analytics Eurographics, 2010, pp.57-86, ISBN: 978-3-905673-77-7 DEFINITIONS VISUAL The creation of tools and techniques to enable people to: Synthesize information and derive insight from massive, dynamic, ambiguous, and often conflicting data. Detect the expected and discover the unexpected. Provide timely, defensible, and understandable assessments. Communicate these assessment electively for action. 2
DEFINITIONS BIG DATA Volume Variety Velocity Variability Complexity it s relative - when an organization s ability to handle, store and analyze data exceeds its current capacity they have big data DEFINITIONS VISUAL P. C. Wong and J. Thomas. Visual analytics. IEEE Computer Graphics and Applications, 24(5):20 21, 2004. Daniel Keim, Jörn Kohlhammer, Geoffrey Ellis and Florian Mansmann (Eds.) Mastering The Information Age Solving Problems with Visual Analytics Eurographics, 2010, pp.57-86, ISBN: 978-3-905673-77-7 3
ENTERPRISE THE LANDSCAPE ENTERPRISE VISUALIZATION FIT FOR PURPOSE 4
VISUALIZATION WHICH OPTION TO PROPOSE? Business Intelligence Personas: General Data: General Purpose: Basic visualization/reporting Value: Easy understanding of information, trends and status Data Explorer Personas: SAS Drug Development User Data: SAS Datasets in SDD Purpose: Quick view, filtering, structure exploration Value: Built in capability to browse and explore data sets JMP Clinical Personas: Scientist, Medical Reviewer, Clinical Researcher Data: SDTM-like data sets Purpose: Clinical Trial specific visualization (e.g. Patient Profile, Patient Narratives, Medical Monitor Dashboard, Fraud Detection) Value: Wide variety of pre-built clinical solutions. Easy to use non-coder analytical viewer Visual Analytics Personas: Senior Management, non statisticians Data: Large/Big data, Aggregated, Mixed purpose/secondary use Purpose: Ad hoc Exploration, Cohort Definition, Outlier Identification, Big Data Exploration Value: Intuitive, nonstatistician explorer, Responsive large data exploration, Visually pleasing reporting and dashboards SAS BI/SAS GRAPH ROBERT ALLISON Clarke Error Grid Plot Hemoglobin Error Grid Flu Tracking Analytics U.S. Fuel Ethanol Jim Goodnight dot plot Election Forecast Map Exposure to Meningitis Interactive Human Body Custom SG Matrix Star Trek's Dashboard Baby Name Popularity China's Favorite Mobile Oil & Gasoline Price Oil & Gas Correlation Basketball Forecasting http://www.robslink.com/sas/home.htm 5
STATISTICAL DISCOVERY FROM SAS VISUAL ADVANCED DATA VISUALIZATION Move from big data to right data Dramatically increase the speed to insight Meaningfully support data driven decision making Intuitive and collaborative access to information Explore data independently No sub-setting or sampling required 6
DELIVERS A SINGLE SOLUTION FOR FASTER, SMARTER DECISIONS Central Entry Point Integration Role-Based Views PREPARE EXPLORE DESIGN Manage data Load and join data Create calculated columns Perform ad-hoc data exploration Insights generated through analytic visualizations Create dashboard style reports for web or mobile DELIVER SAS Mobile BI - native tablet applications delivering interactive reports Web and PDF IN-MEMORY ENGINE VISUAL APPLIED ADVANCED DATA VISUALIZATION Clinical Research Cross study/meta analysis Molecular/genomics Patient Safety Post marketing surveillance Disproportionality analysis Disease Management Monitor large populations Identify cohorts of interest Outcomes Research Hypothesis generation Real time interactive exploration 7
KEY HIGHLIGHTS OF VISUAL 6.1 Analytics Visualization Mobility Enterprise Forecasting Precision Layout New Visualizations Enhanced Security Correlation Analysis Enhanced Interactivity Enhanced Interactivity Advanced Data Preparation Advanced Regression Enhanced Visualizations Enhanced Security Incremental Data Enhanced Box Plot Analysis Custom Calculations Android Support Share Data (Export) ANALYTICAL CAPABILITIES Intelligent Forecasting Model selection Clear explanation of Forecasting results Details in graphical and tabular format Share results. 8
ANALYTICAL CAPABILITIES Multiple regression model options Clear explanation of analytical results Details in graphical and tabular format Share results. ENHANCED VISUALIZATION & INTERACTIVITY Brushing Mode allows for interactivity between multiple visuals Dynamic filtering allows for quickly selecting inclusions / exclusions Share results. 9
REPORTING CAPABILITIES Precision layout of report objects New visualizations New ways to create report interactivity Embed Stored Process results REPORTING CAPABILITIES Precision layout of report objects New visualizations New ways to create report interactivity Embed Stored Process results 10
DATA PREPARATION CAPABILITIES Operations for analytical data preparation Robust data query and structuring Easy to use SQL based operations DATA ADMINISTRATION Table Level Security Row-Level Security 11
DEMONSTRATION 12