3/28/2017 5:03 PM { Data Mining leicht gemacht - der innovative Ansatz für Data Mining von Microsoft im Überblick } Martin Vach Technologieberater.

Präsentation zum Thema: "3/28/2017 5:03 PM { Data Mining leicht gemacht - der innovative Ansatz für Data Mining von Microsoft im Überblick } Martin Vach Technologieberater."—  Präsentation transkript:

1 3/28/2017 5:03 PM { Data Mining leicht gemacht - der innovative Ansatz für Data Mining von Microsoft im Überblick } Martin Vach Technologieberater Business Intelligence Microsoft Deutschland GmbH © 2007 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

2 Agenda Einleitung - Data Mining und prediktive Analyse
3/28/2017 5:03 PM Agenda Einleitung - Data Mining und prediktive Analyse Umfassendes und vollständiges Angebot SQL Server Data Mining-Plattform Data Mining-Add-Ins für Office 2007 Integriertes Data Mining Einbindung und Erweiterbarkeit der Data Mining-Plattform SQL Server 2008 Neuigkeiten im Bereich Data Mining Zusammenfassung © 2007 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

3 Agenda Data Mining und prediktive Analyse 3/28/2017 5:03 PM
© 2007 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

4 Microsoft Solution Center
Was ist Data Mining? “Data Mining ist die halbautomatische Extraktion von Mustern, Änderungen, Assoziationen, Anomalien und anderen statistisch signifikanten Strukturen aus großen Datenmengen.” Robert Grossman Basis sind Methoden und Verfahren aus der Statistik und der künstlichen Intelligenz (KI) Data Mining wird oft als ein Teilgebiet von Business Intelligence betrachtet Abfragen, Reporting, OLAP Data Mining Was geschah? Manuell / Interaktiv / Reaktiv Historische Sicht Warum und wie geschieht etwas? Automatisierte Verfahren Historie, Gegenwart & Zukunft SQL Server Data Mining© Microsoft Corporation. All rights reserved. This presentation is for informational purposes only. Microsoft makes no warranties, express or implied, in this summary.

5 Data - Mining Teilgebiete und Beispiele “Automatisierte Datenanalyse“
Microsoft Solution Center Data - Mining Teilgebiete und Beispiele “Automatisierte Datenanalyse“ Zusammenhänge erkennen (Explorativ) Muster finden Vorhersagen machen (Prediktiv) Analytische CRM: Kundenabwanderungs-Analysen (Churn-Analysis) Kunden-Scoring, Potentialanalysen, Erkennung hochwertiger Kunden Zielgruppen-Marketing: Kampagnen-Optimierung Cross-Selling: Web-Shop Personalisierung, Warenkorbanalyse Aufdeckung von Anomalien und Abweichungen (Schwachstellenanalyse): Entdeckung von Betrugsversuchen (Fraud Detection) Abweichung vom geplanten Verhalten: Prozess- und Produktionsfehler Vorhersage von Risiken („Frühwarnsysteme“): Versicherungs-, Kredit- oder Gesundheitsrisiken SQL Server Data Mining© Microsoft Corporation. All rights reserved. This presentation is for informational purposes only. Microsoft makes no warranties, express or implied, in this summary.

6 Beispiel: Clustering (Segmentierung) Kundensegmentierung, z. B
Beispiel: Clustering (Segmentierung) Kundensegmentierung, z.B. für Zielgruppen-Marketing Einteilung der Daten in homogene Gruppen, wobei die Gruppen sich möglichst stark von einander unterscheiden Alter Männlich Weiblich Sohn Tochter Eltern

7 Beispiel: Entscheidungsbaum Kaufverhalten vorhersagen
Bestimmung der relevanten Einflussgrößen für eine vorherzusagende Größe (Bike Buyer Y/N)

8 Data Mining und prediktive Analyse
Eigenständigkeit der Software Prediktive Analyse Data Mining Pro-Aktiv Interaktiv OLAP Ad-Hoc Reporting Standard-Reporting Passiv Nutzen und Einsicht Präsentation Exploration Erkenntnisse

9 { Von OLAP zu Data Mining }
3/28/2017 5:03 PM { Von OLAP zu Data Mining } Demo Kundenverhalten analysieren und vorhersagen A) OLAP-Analyse B) Aufbau eines einfachen Vorhersage-Modells © 2007 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

10 SQL Server 2008 Data Mining Bestandteil der SQL Server 2008 Analysis Services
Complete Pervasive Delivery through Microsoft Office Comprehensive Development Environment Enterprise Grade Capabilities Rich and Innovative Algorithms Integrated Native Reporting Integration In-Flight Mining during Data Integration Insightful Analysis Predictive KPIs Extensible Predictive Programming Custom Algorithms and Visualizations

11 Agenda Umfassendes und vollständiges Angebot
3/28/2017 5:03 PM Agenda Umfassendes und vollständiges Angebot SQL Server Data Mining-Plattform © 2007 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

12 Vollständige Data Mining-Plattform Erfüllung aller relevanten Anforderungen
Rapid Development High Availability Superior Performance and Scalability Robust Security Features Enhanced Manageability Analysis Services

13 Komfortable Entwicklungsumgebung BI Development Studio
Intuitiver Data Mining Wizard Grafischer Data Mining Designer visuelle & statistische Validierung Klassifikations-Matrizen Lift-Charts Profit-Charts Kreuz-Validierung Effizienter Zugriff auf die Quelldaten Caching Filter Aliasing

14 { Vorhersagemodell Kaufverhalten, Teil 2}
3/28/2017 5:03 PM { Vorhersagemodell Kaufverhalten, Teil 2} Demo - Modell-Validierung - Vorhersage-Query © 2007 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

15 Umfangreiches Angebot an Algorithmen
Innovative Algorithmen von Microsoft Research Standard- Algorithmen wie ARIMA Breites Spektrum von Möglichkeiten zur Erstellung optimaler Modelle Algorithmen zur Lösung typischer Geschäftsprobleme Daten - Exploration Warenkorbanalyse Abwanderungsanalyse Kundensegment Analysen Zeitreihenanlayse (Forecast) Unsupervised Learning Webseiten-Analyse Kampagnen-Analyse Daten-Qualitäts-Fragen Text-Analyse/Text Mining

16 Vollständiger Satz von Algorithmen
Microsoft Solution Center Vollständiger Satz von Algorithmen Time Series Decision Trees Clustering Sequence Clustering Association Naive Bayes Neural Net Linear Regression Logistic Regression SQL Server Data Mining© Microsoft Corporation. All rights reserved. This presentation is for informational purposes only. Microsoft makes no warranties, express or implied, in this summary.

17 Data Mining - Aufgaben und Algorithmen
Task Description Algorithms Market Basket Analysis Discover items sold together to create recommendations on-the-fly and to determine how product placement can directly contribute to your bottom line. Association Decision Trees Churn Analysis Anticipate customers who may be considering canceling their service and identify the benefits that will keep them from leaving. Linear Regression Logistic Regression Market Analysis Define market segments by automatically grouping similar customers together. Use these segments to seek profitable customers. Clustering Sequence Clustering Forecasting Predict sales and inventory amounts and learn how they are interrelated to foresee bottlenecks and improve performance. Time Series Data Exploration Analyze profitability across customers, or compare customers that prefer different brands of the same product to discover new opportunities. Neural Network Unsupervised Learning Identify previously unknown relationships between various elements of your business to inform your decisions. Web Site Analysis Understand how people use your Web site and group similar usage patterns to offer a better experience. Campaign Analysis Spend marketing funds more effectively by targeting the customers most likely to respond to a promotion. Naïve Bayes Information Quality Identify and handle anomalies during data entry or data loading to improve the quality of information. Text Analysis Analyze feedback to find common themes and trends that concern your customers or employees, informing decisions with unstructured input. Text Mining

18 Agenda Umfassendes und vollständiges Angebot
3/28/2017 5:03 PM Agenda Umfassendes und vollständiges Angebot Data Mining-Addins für Office 2007 © 2007 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

19 Data Mining-Addins für Office 2007
3/28/2017 5:03 PM Data Mining-Addins für Office 2007 Table Analysis Tools for Excel 2007 Data Mining Client for Excel 2007 Data Mining Template for Visio 2007 Kostenlose Add-Ins (Download) für Office Excel 2007, Office Visio 2007 Voraussetzung: Server mit SQL 2005 SP2 / SQL 2008 Neue Version für SQL Server 2008 verfügbar mit SQL 2008 RTM MICROSOFT CONFIDENTIAL © 2006 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

20 Der Data Mining Prozess und Excel Vorgehensmodell "CRISP-DM"
Verstehen der fachlichen Fragestellung Verstehen der Datenbasis Aufbereitung der Daten Data Bereitstellung und Nutzung Modellierung Validierung

21 Define Data Identify Task Get Results
Data Mining Add-Ins für Microsoft Office 2007 Die “DIG” – Formel: Data Mining am Arbeitsplatz Define Data Identify Task Get Results “What Microsoft has done is to make data mining available on the desktop to everyone” - David Norris, Associate Analyst, Bloor Research

22 Microsoft Solution Center
Data Mining mit Office 2007 Tabellenanalyse-Tool (Table Analysis) für Excel 2007 – Leicht verwendbare Assistenten für einfache Data Mining-Aufgaben Data Mining Client für Excel 2007 – Vollständiger Entwicklungszyklus für alle Data Mining Schritte: Daten-Aufbereitung Modelle erstellen, auch mit Excel-Daten Testen und Validieren der Modelle Modelle untersuchen Modelle verwalten Vorhersagen machen, auch mit Excel-Daten Data Mining Vorlagen für Visio – Darstellung von Data Mining Modellen als Visio-Objekte SQL Server Data Mining© Microsoft Corporation. All rights reserved. This presentation is for informational purposes only. Microsoft makes no warranties, express or implied, in this summary.

23 { Data Mining mit Excel 2007 }
3/28/2017 5:03 PM { Data Mining mit Excel 2007 } Demo Table Analysis Add-In - Key Influencer Data Mining Add-In - Vorhersagemodell © 2007 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

24 demo { Predict Outcomes } Forecasting with the Time Series algorithm
Using the Prediction Calculator © 2008 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

25 Office-SQL Server Data Mining Architektur
3/28/2017 5:03 PM SQL Server Datenbank (oder andere DBs) SQL Server Analysis Services Excel 2007 Add-ins Mining Models Modeling Query Excel Data Data Data Source Server (optional) Client Server MICROSOFT CONFIDENTIAL © 2006 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

26 Data Mining Add-Ins für Microsoft Office 2007 Unterstützung des gesamten Data Mining-Prozess
Data Preparation Explore, clean and set up your data for data mining Data Modeling Build patterns and trends from data to make predictions Accuracy and Validation Test and validate your model Model Usage & Management Browse, modify, and manage existing mining models that are stored on an instance of Analysis Services Documentation Trace your actions as Data Mining Extensions (DMX) statements or as Analysis Services Scripting Language (ASSL).

27 Agenda Integriertes Data Mining 3/28/2017 5:03 PM
© 2007 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

28 Data Mining und Reporting Services
Erstellung von Berichten mit Vorhersagen mittels Data Mining-Abfragen als Datenquelle Query-Builder für DMX- Abfragen im Report Designer verfügbar Entwurf von Parameter- getriebenen Berichten auf Basis von Vorhersage- Wahrscheinlichkeiten Z.B. Anzeige von Risiko- Kunden mit Abwanderungs- Wahrscheinlichkeit > 65%

29 Nutzung von Data Mining im ETL-Prozess Erweiterung der Möglichkeiten von SSIS
Lösung von ETL-Aufgaben Markierung anormaler Daten Klassifizierung von Kunden oder Geschäftsobjekten Erkennung fehlender Werte Aufbereitung von unstrukturierten Daten mittels Text Mining ETL-Erweiterungen Scoring (z.B. von Kunden) mittels der DM-Query-Task Trainieren von DM-Modellen mittels der DM-Training- Destination

30 Data Mining und OLAP-Cubes
Nutzung von OLAP- Cubes als Quelle für Data Mining Einbinden von DM- Ergebnissen als OLAP- Dimension, z.B. Kundencluster Nutzung von Vorhersage-Funktionen in MDX-Berechnungen und für KPI’s

31 Data Mining und Performance Management Integration mit PerformancePoint Server
Nutzung der Zeitreihen- Analyse als “Analytical View” in Dashboards Kombination von prediktiven und historischen KPI’s für aussagekräftigere Dashboards Vorhersage der zukünftigen Ergebnisse im Vergleich zu den Zielen zur Erkennung möglicher Herausforderungen Analyse und Monitoring von Trends bei den Haupteinflussgrößen

32 Agenda Einbindung und Erweiterbarkeit der Data Mining-Plattform
3/28/2017 5:03 PM Agenda Einbindung und Erweiterbarkeit der Data Mining-Plattform © 2007 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

33 Erweiterbarkeit – Data Mining API’s
Plug-in Algorithms Add custom data mining algorithms Visualizations Redistributable Viewer - embed standard visualizations in your application Plug-in Viewer APIs - embed custom visualizations in your application PMML Exchange models with other software vendors XMLA Industry standard metadata Data mining Extensions (DMX) SQL-like query language ADOMD.NET and OLE DB Access and query models from clients or stored procedures AMO Management interfaces Erweiterungen Einbindung

34 Nutzung von Data Mining in Anwendungen „Predictive Programing“
Einbettung von Data Mining Integration einer “Recommendation Engine” Aktualisierung von Modellen auf Basis der aktuellsten Daten on the fly Ausreißer-Erkennung on the fly z.B. bei Datenerfassung Muster-Erkennung Anzeige wesentlicher Indikatoren für Metriken Erkennung von Profilen für Abwanderungen oder hochwertige Kunden Vorhersage Empfehlung relevanter Produkte Darstellung von Risiko-Kunden bzw. Abwanderungs-Wahrscheinlichkeit Optimierung von Promotions- Kampagnen für Kunden mit hohem Wert (life time value) Einbindung von Data Mining in Business-Anwendungen mittels komfortabler API’s ?

35 Vorhersagen mit DM-Modellen
Microsoft Solution Center Vorhersagen mit DM-Modellen DMX Data Mining Extensions SQL ähnliche Sprache für Erstellung Abfrage von DM Modellen DM-Funktionen Predict(), PredictProbability, CaseLikelihood, etc User-defined functions, Parametrisierte Abfragen DMX Prediction Joins für Vorhersagen SELECT t.ID, CPModel.Plan FROM CPModel PREDICTION JOIN OPENQUERY(…,‘SELECT * FROM NewStudents’) AS t ON CPModel.Gender = t.Gender AND CPModel.IQ = t.IQ DM-CPModell Gender IQ Plan ID Gender IQ NewStudents SQL Server Data Mining© Microsoft Corporation. All rights reserved. This presentation is for informational purposes only. Microsoft makes no warranties, express or implied, in this summary.

36 Agenda SQL Server 2008 Neuigkeiten im Bereich Data Mining
3/28/2017 5:03 PM Agenda SQL Server 2008 Neuigkeiten im Bereich Data Mining © 2007 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

37 SQL Server 2008 Data Mining Neuigkeiten im Überblick
Verbesserung der Engine und der Algorithmen Anforderungen vieler professioneller DM-Kunden Verbesserung im Bereich Mining-Strukturen BI Development Studio, Handhabung, Aufwand Data Mining AddIns für Office 2007 Durchgängige und komfortable Benutzerführung Warenkorbanalyse Prediction Calculator © 2008 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

38 SQL Server 2008 - Neuigkeiten Verbesserung im Bereich Zeitreihen
Basis ist der ARTXP Zeitreihen-Vorhersage-Algorithmus Entwickelt von MS Research Adressiert ein Kern-Problem: suche die bestmögliche Vorhersage für den nächsten Schritt in der Serie Weniger geeignet für Langzeit-Vorhersagen SQL Server 2008 ARTXP nach wie vor verfügbar Optimal für kurzfristige Vrohersagen Zusätzlich verfügbar: ARIMA Der bekannteste und verbreiteste Zeitreihen-Algorithmus Gute Kenntnisse bei praktisch allen Data Mining-Experten Aktzeptable Vorhersagen bei Projektion auf mehr als 10 Schritte

39 SQL Server 2008 - Neuigkeiten Verbesserungen im Bereich Mining Strukturen
Aufteilung in Trainings- und Test-Partitionen Automatisch, manuell oder programmatisch Abfragen gegen Struktur-Cases und Struktur-Spalten Ermöglicht Drillthrough aus einem Cluster-Modell um zusätzliche Daten anzuzeigen, die nicht im Modell benutzt werden (z.B. eine Mail-Adresse) Filterung von Daten beim Aufbau von Modellen Beispiel: Erstelle getrennte Modelle für männliche und weibliche Kunden Erstellung nicht-kompatibler Modelle in derselben Struktur Modelle, die die kontinuierliche und die diskretisierte Version derselben Spalte nutzen, können dieselben Struktur nutzen Kreuz-Validierung (Cross-validation) Erleichtert Verstehen der Modell-Genauigkeit bzw. Modell-Güte Automatischer Test des Modell gegen mehrere Subsets von Trainingsdaten und Vergleich der Ergebnisse

40 { Weitere Data Mining Beispiele (if we have time…) }
3/28/2017 5:03 PM { Weitere Data Mining Beispiele (if we have time…) } Demo Table Analysis - Prediction Calculator Kreuz-Validierung © 2007 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

41 Enhanced Mining Structures
Split data into training and testing partitions more effectively Query against structure data to present complete information beyond the scope of the model Build models over filtered data Create incompatible models within the same structure Use cross-validation to: Test multiple models simultaneously Confirm the stability of results given more or less data Better Time Series Support Accuracy & Stability Combine best of both worlds blending ARTXP for optimized near-term predictions and ARIMA for stable long term predictions Prediction Flexibility Build a forecasting model on one series and apply the patterns to data from another series. What If Anticipate the impact of changes in near-term future values, on long-term forecasts More Data Mining Add-Ins for Office 2007 New Analysis Tools Generate interactive forms for scoring new cases with Prediction Calculator Discover the relationship between items, which are frequently purchased together with Shopping Basket Analysis New Query and Validation Tools Choose training and test sets from mining structures Render richly-formatted cross validation and accuracy reports in Excel Leverage model documentation for reference and collaboration

42 Agenda Zusammenfassung 3/28/2017 5:03 PM
© 2007 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

43 ABS-CBN Interactive (ABSi)
Subsidiary of the largest integrated media and entertainment company in the Philippines Wireless Services Firm Doubles Response Rates with SQL Server 2005 Data Mining Challenge Selling custom ring tones and other downloadable content for mobile phone users requires staying in tune with the market. Searching transactional data for hints on what to offer users in cross-selling value-added mobile services took days and didn’t provide customer-specific recommendations. Solution ABSi deployed Microsoft® SQL Server™ 2005 to use its data mining feature to determine product recommendations. Benefit More accurate and personalized service recommendations to customers Doubling response rates from marketing campaigns Ad hoc reporting in minutes, not days Eight times faster data mining process Faster data mining prediction “Our management is very impressed that we could double our response rate through our SQL Server 2005 data mining … managers of other services ask us to provide the same magic for them—which is what we will do with the full project rollout” - Grace Cunanan, Technical Specialist, ABS-CBN Interactive

44 Data Mining Kundenbeispiele
.8 TB SS2005 DW for Ring-Tone Marketing Uses Relational, OLAP and Data Mining 5 TB DW, serving the 2nd largest global HMO with over 3000 OLAP users. Developed data mining solution to identify members who would most benefit from proactive intervention to prevent health deterioration. 3 TB end-to-end BI decision support system Oracle competitive win End-to end DW on SQL Server, including OLAP Extensive use of Data Mining Decision Trees 1.2 TB, 20 billion records Large Brazilian Grocery Chain .88 TB DW at main TV network in Italy Increased viewership by understanding trends .5 TB DW at US Cable company End to end BI, Analysis and Reporting

45 Zusammenfassung Data Mining mit Microsoft
3/28/2017 5:03 PM Einfacher Zugang – Erhöhung der Reichweite Data Mining für jeden „Knowledge Worker“ Jede Anwendung kann Data Mining nutzen durch ADOMD.NET - ohne komplexe API‘s Anwender und Entwickler nutzen gewohnte Umgebung Kosten und Nutzen Attraktives Lizenzmodell - kein Lizenzkosten-KO mehr Geringe Einstiegskosten - Schnelle Ergebnisse Kein Data Mining mehr im „Elfenbeinturm“ Vollständig – Integriert - Erweiterbar © 2007 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

46 Summary Complete Integrated Extensible
Pervasive Delivery through Microsoft Office empowers all users with predictive insight Comprehensive Development Environment delivers an intuitive and rich environment Enterprise Grade Capabilities provide enhanced server advantages Rich and Innovative Algorithms support common business problems effectively Integrated Native Reporting Integration seamlessly infuses prediction into reports In-Flight Mining during Data Integration dynamically enhances data quality & relevance Insightful Analysis enables to slice data by the hidden patterns within Predictive KPIs extend monitoring with insights to future performance Extensible Predictive Programming embeds prediction within the application Custom Algorithms & Visualizations provide the flexibility to meet uncommon needs

47 Ask the Experts Martin.Vach@microsoft.com
Wir freuen uns auf Ihre Fragen: Technische Experten stehen Ihnen während der gesamten Veranstaltung in der Haupthalle zur Verfügung.

48 Weitere Informationen
ACHTUNG: Februar CTP6 Version ist seit heute verfügbar

49 Danke für Ihre Aufmerksamkeit !
3/28/2017 5:03 PM Danke für Ihre Aufmerksamkeit ! © 2007 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

50 3/28/2017 5:03 PM © 2007 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION. © 2007 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

51 Anhang 1 How to Build a Successful Predictive Analysis Project?

52 What’s Your Problem? Success needs… Right problem Right criteria
Right data Right results Right delivery

53 Is Yours a Data Mining Problem?
Right Problem Is Yours a Data Mining Problem? Driven by data, or by business rules? Driven by what you know, or what you don’t know? Traditional BI Predictive Analysis Exploration Discovery Drill down Classification Trending Prediction Force constraints Discover outliers Apply rules & models Find patterns & relationships

54 The Top 5 Business Scenarios
Right Problem The Top 5 Business Scenarios Cross-sell and up-sell Campaign management Customer acquisition Budget and forecasting Customer retention

55 Scenarios, Tasks and Techniques
Right Problem Scenarios, Tasks and Techniques Define scenarios by Data Mining tasks to be performed Classification Estimation Prediction Association Clustering Define tasks by techniques to be used The algorithms used to build models

56 From Scenarios to Tasks
Right Problem From Scenarios to Tasks Scenario Tasks Cross-sell and Up-sell Association Campaign management Classification, Clustering Customer acquisition Clustering Budget and forecasting Prediction Customer retention Classification, Estimation

57 From Tasks to Techniques
Right Problem From Tasks to Techniques Tasks Techniques Association Association rules, Decision trees Classification Decision Trees, Neural Net, Naïve Bayes Clustering Estimation Logistic, Linear Regression Prediction Time Series

58 Criteria: Are You Just Fishing?
Right Criteria Criteria: Are You Just Fishing? How will you measure success? Technical: Lift and accuracy Business: ROI and KPIs Set good criteria Relevant and actionable Strategic vs. operational relevance Realistic and achievable

59 Your Biggest Job Data exploration and preparation Don’t forget GI-GO
Right Data Your Biggest Job Data exploration and preparation Don’t forget GI-GO Is your data… Complete, accurate, timely, typical? Get to know Integration Services Merge and transform source data Filter and sample

60 Prepare Your Data Clean Clarify Simplify
Right Data Prepare Your Data Clean Remove duplicates and missing values Remove out-of-range data Clarify True, False and NULL boolean fields Remove synonyms UnitPrice * Qty = TotalPrice Calculate derived values Simplify Bucket or Group continuous or many-valued columns Age, Profession

61 We All Crave Some Validation…
Right Results We All Crave Some Validation… Technical Evaluation Use lift charts & classification matrices Compare training and test sets Business Evaluation Test results against business metrics Review regularly Don’t take trust for granted

62 Delivering the Results
Right Delivery Delivering the Results Aim for a seamless experience Enhance existing information Repurpose existing skills Win your users

63 Choosing the User Environment
Right Delivery Choosing the User Environment Sales and operations Not software specialists Can be overwhelmed by detail or ambiguity Information workers Explorers, mostly live in Microsoft Office Excel BI Analysts Love their BI tools of choice Executives Dashboards, Scorecards Also overwhelmed by detail or ambiguity

64 3/28/2017 5:03 PM © 2007 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION. © 2007 Microsoft Corporation. All rights reserved. Microsoft, Windows, Windows Vista and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.