SAS Analytics Day An Introduction to SAS Enterprise Miner and SAS Forecast Server André de Waal, Ph.D. Analytical Consultant
Agenda 1. Introduction to SAS Enterprise Miner 2. Basics 3. Enterprise Miner Demo 4. Introduction to SAS Forecast Server 5. Basics 6. Forecast Server Demo
Agenda 1. Introduction to SAS Enterprise Miner 2. Basics 3. Enterprise Miner Demo 4. Introduction to SAS Forecast Server 5. Basics 6. Forecast Server Demo
Introduction to Enterprise Miner SAS Enterprise Miner streamlines the data mining processto create highly accurate predictive and descriptive models based on vast amounts of data gathered from across an organization.
Introduction to Enterprise Miner Powerful, easy-to-use GUI, as well as batch processing for large jobs Data preparation, summarization and exploration Advanced predictive and descriptive modeling Fast, easy and self-sufficient way for business users to generate models Business-based model comparisons, reporting and management Automated scoring process Open, extensible design & scalable processing
Introduction to Enterprise Miner Benefits Support the entire data mining process with a broad set of tools. Regardless of your data mining preference or skill level, SAS provides flexible software that addresses complex problems. Going from raw data to accurate, business-driven data mining models becomes a seamless process, enabling the statistical modeling group, business managers and the IT department to collaborate more efficiently. Build better models with a versatile data mining workbench. SAS Enterprise Miner includes an interactive self-documenting process flow diagram environment that dramatically shortens model development time for statisticians and data miners. It efficiently maps the entire data mining process to achieve the best possible results.
Introduction to Enterprise Miner Enable business analysts and subject-matter experts with limited statistical skills to generate predictive models for a variety of business scenarios. TheSAS Rapid Predictive Modelertask running from SAS Enterprise Guide or the SAS Add-In for Microsoft Office (Excel only) enables business users to automatically generate predictive models and act on them quickly and effectively. Analytic results can be consumed in simple and easy-tounderstand charts to derive the insights needed for better decision making.
Introduction to Enterprise Miner Enhance accuracy of predictions and easily share reliable information to improve the quality of decisions. Better-performing models with new innovative algorithms enhance the stability and accuracy of predictions, which can be verified easily by visual model assessment and validation metrics. Both analytical and business users enjoy a common, easy-to-interpret visual view of the data mining process. Predictive results and assessment statistics from models built with different approaches can be displayed side-by-side for easy comparison. The resulting diagrams serve as selfdocumenting templates that can be updated easily or applied to new problems without starting over.
Introduction to Enterprise Miner Ease the model deployment and scoring process. Scoring the process of applying a model to new data is the end result of many data mining endeavors. SAS Enterprise Miner automates the tedious scoring process and supplies complete scoring code for all stages of model development in SAS, C, Java and PMML. The scoring code can be deployed in a variety of real-time or batch environments within SAS, on the Web or directly in relational databases. The outcome is faster implementation of data mining results.
SAS Enterprise Miner Interface Tour 10
SAS Enterprise Miner Interface Tour Menu bar and shortcut buttons 11
SAS Enterprise Miner Interface Tour Project panel 12
SAS Enterprise Miner Interface Tour Properties panel 13
SAS Enterprise Miner Interface Tour Help panel 14
SAS Enterprise Miner Interface Tour Diagram workspace 15
SAS Enterprise Miner Interface Tour Process flow 16
SAS Enterprise Miner Interface Tour Node 17
SAS Enterprise Miner Interface Tour SEMMA tools palette 18
The Analytic Workflow Analytic workflow Define analytic objective Select cases Extract input data Validate input data Repair input data Transform input data Apply analysis Generate deployment methods Integrate deployment Gather results Assess observed results Refine analytic objective 19 19
Enterprise Miner Demo Ozone data set The goal is to predict the Ozone concentrations in the air given 9 exploratory variables: Wind speed (mph) Inversion base temperature (degrees F) Visibility (miles) Calendar day, between 1 and 366 Etc.
Introduction to Forecast Server SAS Forecast Server is our flagship forecasting software and can automatically generate large quantities of statistically based forecasts without the need for human intervention, unless so desired. It operates in either batch mode or interactively through the Forecast Studio GUI. SAS Forecast Server automatically chooses the most appropriate forecasting model from an extensible model repository, optimizes model parameters, and generates the forecasts. Forecasters can focus attention on problematic or high-value forecasts, and conduct what-if analysis through the Scenario Analyzer tool.
Introduction to Forecast Server Easy-to-use GUI Large-scale automatic forecasting Easy manageability
Introduction to Forecast Server Motivation: The Large-Scale Forecasting
Introduction to Forecast Server A skilled analyst can forecast a single time series by applying good judgment based on knowledge and experience using various proven time-series analysis techniques utilizing good software based on sound statistical theory.
Introduction to Forecast Server Modern businesses require efficient, reliable forecasts for many series. These forecasts usually need to be updated on a regular basis. There are not sufficient resources to apply the single series forecasting approach to all series that need to be forecast. The series might be hierarchically arranged and require reconciliation of forecasts at different levels.
Introduction to Forecast Server Large-Scale Forecasting Scenario 80% can be forecast automatically. 10% requires extra effort. 10% cannot be forecast accurately. Time Series Data
Introduction to Forecast Server Benefits Delivers forecasts in a quick and timely manner through a userfriendly graphical interface. Provides forecasts that reflect the realities of the business, improving your ability to plan future events with confidence. Improves forecasting performance across all products and locations, at any level of aggregation.
SAS Forecast Studio Interface Tour 28
SAS Forecast Studio Interface Tour 29
SAS Forecast Studio Interface Tour 30
SAS Forecast Studio Interface Tour Menu Bar and Shortcut Buttons 31
SAS Forecast Studio Interface Tour Menu Bar and Shortcut Buttons 32
SAS Forecast Studio Interface Tour The Active Series Overview Panel 33
SAS Forecast Studio Interface Tour The Four View Tabs 34
SAS Forecast Studio Interface Tour The Forecasting View 35
SAS Forecast Studio Interface Tour The Modeling View 36
SAS Forecast Studio Interface Tour The Modeling View MSL The Model Selection List (MSL) associated with the highlighted series 37
SAS Forecast Studio Interface Tour The Scenario Analysis View 38
SAS Forecast Studio Interface Tour The Series View 39
Disseminate observed results Refine forecasting objectives Accommodate data updates The Forecasting Workflow Forecasting workflow 40 Define forecasting objectives Select series, specify the data hierarchy Validate series and hierarchy Repair series Accumulate series, create the hierarchy Apply analysis and generate forecasts Reconcile forecasts Gather results Assess observed results Apply forecast overrides
Forecast Server Demo Toys data set The goal is to forecast the number of toys that will be sold in a week Wine data set The goal is to predict wine demand over four wine types and four distribution regions
For more details contact presenter at: Presenter: André de Waal Phone: 1-919-531-6575 Email: andre.dewaal@sas.com