Programma

With data increasingly being seen as a critical corporate asset, more organisations are embracing the concepts and practices of Data Governance. As a result Data Governance is today one of the hottest topics in data management, focusing both on how Governance driven change can enable companies to gain better leverage from their data through enhanced Business Intelligence, Data Analytics and so on, and also to help them design and enforce the controls needed to ensure they remain compliant with increasingly stringent laws and regulations, such as GDPR.

Despite this rapidly growing focus, many Data Governance initiatives fail to meet their goals, with only around one in five fully achieving expectations. Why is the failure rate so high? There are many factors, but one key reason is that implementing Data Governance without aligning it with a defined enterprise and data architecture is fraught with dangers. Linking Architecture with data accountability, a core principle of Data Governance, is essential.
This session will outline why Data Governance and Architecture should be connected, how to make it happen, and what part Business Intelligence and Data Warehousing plays in defining a robust and sustainable Governance programme.

This talk will cover:
• What is Data Governance and what it is not
• Key reasons for Data Governance failure & disappointment
• The key components of enterprise architecture – Business, Process and Data
• The synergies between architecture and Governance – how do they reinforce each other?
• How artefacts from both disciplines can be combined and applied to ensure success
• The implications for Business Intelligence and Data Warehousing
• Several use cases of successes and lessons learned

In de complexe wereld van halfgeleider productie worden dagelijks enorme hoeveelheden zeer gevarieerde data gegenereerd. ASML, wereldleider op het gebied van machines voor de productie van halfgeleiders, implementeert een central data lake waarin data verzameld wordt in een centrale omgeving en die vanuit daar beschikbaar gesteld wordt voor rapportage en analyse. Dit central data lake bevat ook een zogenaamd analytics lab voor gedetailleerde exploratie van data en het faciliteren van data science toepassingen. Het beheersen van deze snel veranderende data is een enorme uitdaging. In deze sessie bespreken we aan de hand van een aantal voorbeelden ASML’s aanpak voor de volgende uitdagingen:

• Hoe kunnen gebruikers, analisten en data scientists informatie in het central data lake ontdekken en vinden zonder te verdrinken in de hoeveelheid en complexiteit van gegevens?
• Hoe borgen we dat gebruikers van het central data lake de data begrijpen en weten waar de data vandaan komt (data lineage)?
• Hoe zorgen we er voor dat de data in het central data lake vertrouwd kan worden (data kwaliteit)?
• Hoe kunnen we als ASML borgen dat toegang tot data in lijn is met eisen van klanten en andere belanghebbenden?
• Welke toegevoegde waarde kan het analytics lab (nog meer) brengen?

When it comes to data analytics, you don’t want to know “how the sausage is made.” The state of most data analytics pipelines is deplorable. There are too many steps; too little automation and orchestration; minimal reuse of code and data; and a lack of coordination between stakeholders in business, IT, and operations. The result is poor quality data delivered too late to meet business needs.

DataOps is an emerging approach for building data pipelines and solutions. This session will explore trends in DataOps adoption, challenges that organizations face in implementing DataOps, and best practices in building modern data pipelines. It will examine how leading-edge organizations are using DataOps to increase agility, reduce cycle times, and minimize data defects, giving developers and business users greater confidence in analytic output.

You Will Learn:
• What is DataOps and why you need it
• The dimensions of DataOps
• The state of DataOps adoption
• DataOps best practices and challenges

Conventional data governance practices come from a simpler time when data management was free from many of today’s challenges, such as self-service reporting and analytics. Traditional data governance focuses on enforcement of controls and gates. While controls and gates continue to be needed, these methods must be complemented with support for the autonomy and agility of the self-service world. Enforcement works together with prevention. Guides and guardrails reduce the need for gating. The need to exercise controls is minimized when curating, coaching, crowdsourcing, and collaboration are integral parts of governance processes. In a self-service world, every data stakeholder plays a part in data governance. In a world where personal data collection is accelerating and algorithms are abundant, ethical guidance is a critical component and an area where governance must take the lead.

You Will Learn:
• Where governance fits within modern data ecosystems, from point of ingestion to reporting and analysis
• How various technologies support governance through the ecosystem
• Process challenges for governing self-service; supplementing controls with collaboration and crowd sourcing
• Engagement models for governing self-service
• Organizational challenges for governing self-service; moving from data stewards to stewardship, curation, and coaching
• Operational challenges for governing self-service; implementing a combination of gates, guardrails, and guides
• The challenges of data ethics and the role of data governance in meeting those challenges

Supervised learning lost moderne analytische uitdagingen op en ondersteunt geïnformeerde besluitvorming. Alhoewel het voorspellend vermogen van machinelearning modellen indrukwekkend kan zijn, moet er wel een actie aan gekoppeld zijn om er profijt van te hebben. Modellen moeten daarnaast ook automatisch worden uitgerold om besluitvorming continue te kunnen ondersteunen en meetbare verschillen te realiseren. En hoewel ook unsupervised learning methoden krachtige analytische mogelijkheden bieden is hier vaak nog geen duidelijke route naar productie. Deze cursus laat zien wanneer welke vorm van machinelearning het beste past bij de business doelstellingen en hoe u meerwaarde kunt behalen uit beide benaderingen.

Regressieanalyse, decision trees, neurale netwerken, samen met vele andere supervised learning technieken, realiseren krachtige voorspellende inzichten wanneer historische resultaatwaarden beschikbaar zijn. Zodra deze modellen gebouwd zijn, genereren supervised learning modellen een score die gebruikt kan worden om automatische besluitvorming in organisaties te ondersteunen. Wij zullen verkennen hoe deze bewegende panelen strategisch gezien bij elkaar kunnen worden gebracht.

Unsupervised methoden zoals clusteranalyse, anomaly detectie en associatieregels zijn verkennend van aard en genereren niet zoals supervised learning modellen een voorspellende score. De vraag is hoe deze modellen in staat kunnen worden gesteld om organisatorische besluitvorming te ondersteunen. Deze cursus zal dat laten zien.

Deze cursus laat een verscheidenheid aan voorbeelden zien, te beginnen met het verkennen en interpreteren van modellen en hun toepassing. Mogelijkheden om met de resultaten van deze modellen verder te gaan zullen worden bezien. U zult ook zien hoe een verzameling van modellen, waaronder businessregels, supervised modellen en unsupervised modellen gezamenlijk kunnen worden toegepast in concrete situaties, zoals bij fraudedetectie en het verstrekken van verzekeringen.

Leerdoelen

• Wanneer moet u supervised en wanneer unsupervised modellen toepassen?
• Mogelijkheden om machinelearning in te zetten voor besluitvorming in uw organisatie
• Hoe kunt u verschillende modellen inzetten voor benaderingen en classificaties in de werkelijkheid?
• Effectieve technieken om resultaten van unsupervised learning toe te passen
• Het interpreteren en monitoren van uw modellen voor continue verbeteringen
• Hoe combineert u op creatieve wijze supervised en unsupervised modellen om tot betere resultaten te komen?

Bestemd voor ú
Deze interactieve workshop is opgezet voor Analisten, Data scientists, IT Professionals, BI Professionals, Technology Planners, Consultants, Business analisten en Projectleiders van analyse opdrachten.

Onderwerpen

1. Model Development Introduction

Current Trends in AI, Machine Learning and Predictive Analytics

• Algorithms in the News: Deep Learning
• The Modeling Software Landscape
• The Rise of R and Python: The Impact on Modeling and Deployment
• Do I Need to Know About Statistics to Build Predictive Models?

2. Strategic and Tactical Considerations in Binary Classification

• What’s is an Algorithm?
• Is a “Black Box” Algorithm an Option for Me?
• Issues Unique to Classification Problems
• Why Classification Projects are So Common
• Why are there so many Algorithms?

3. Data Preparation for Supervised Models

• Data Preparation Law
• Integrate Data Subtasks
• Aggregations: Numerous Options
• Restructure: Numerous Options
• Data Construction
• Ratios and Deltas
• Date Math
• Extract Subtask

4. The Tasks of the Model Phase

• Optimizing Data for Different Algorithms
• Model Assessment
  • Evaluate Model Results
    • Check Plausibility
    • Check Reliability
    • Model Accuracy and Stability
    • Lift and Gains Charts
    • Modeling Demonstration
  • Assess Model Viability
  • Select Final Models
  • Why Accuracy and Stability are Not Enough
  • What to Look for in Model Performance
  • Exercise Breakout Session
    • Select Final Models
    • Create & Document Modeling Plan
    • Determine Readiness for Deployment
    • What are Potential Deployment Challenges for Each Candidate Model?

5. What is Unsupervised Learning?

• Clustering
• Association Rules
• Why most organizations utilize unsupervised methods poorly
  • Case Study 1: Finding a new opportunity
  • Case Studies 2, 3, and 4: How do supervised and unsupervised work together
  • Exercise Breakout Session: Pick the right approach for each case study
  • Data Preparation for Unsupervised
    • The importance of standardization
    • Running an analysis directly on transactional data
  • Unsupervised Algorithms:
    • Hierarchical Clustering
    • K-means
    • Self-Organizing Maps
    • K Nearest Neighbors
    • Association Rules
  • Interpreting Unsupervised
    • Exercise Breakout Session: Which value of K is best?
    • Choosing the right level of granularity
    • Reporting unsupervised results

6. Wrap-up and Next Steps

• Supplementary Materials and Resources
• Conferences and Communities
• Get Started on a Project!
• Options for Implementation Oversight and Collaborative Development

Cloud data warehousing helps to meet the challenges of legacy data warehouses that struggle to keep up with growing data volumes, changing service level expectations, and the need to integrate structured warehouse data with unstructured data in a data lake. Cloud data warehousing provides many benefits, but cloud migration isn’t fast and easy. Migrating an existing data warehouse to the cloud is a complex process of moving schema, data, and ETL. The complexity increases when architectural modernization, restructuring of database schema or rebuilding of data pipelines is needed.

This session describes the benefits, techniques, and challenges of migrating an existing data warehouse to the cloud. We’ll overview the major cloud data warehouse providers and examine some technologies that can ease the pain of cloud migration. Migration pragmatics are described within the framework of a step-by-step approach.

Moving an existing data warehouse to the cloud is not quick, and it isn’t easy, but there are real and substantial benefits. Most companies avoid the “big bang” and migrate incrementally. The techniques and technologies described here will help you to start quickly and navigate through the migration roadmap smoothly.

This is a participative workshop. Participants should arrive with knowledge of one or more on-premises data warehouses that they are considering moving to the cloud. You’ll perform exercises to analyze the upside and downside of cloud migration, to make migration decisions, and to develop a migration plan.

Learning objectives

• The what and why of cloud data warehousing
• The benefits of cloud data warehousing
• The challenges of cloud data warehousing
• Cautions about cloud migration
• How to perform migration analysis and make migration decisions
• A step-by-step approach to data warehouse migration
• Considerations for incremental migration
• The cloud data warehousing technology landscape
• The cloud migration technology landscape
• How to develop a cloud migration plan

Who is It For?

Data management and analytics program and project managers, Data management and analytics architects, Technology strategists and planners, Data engineers and data integration specialists, Anyone with a role in migrating legacy data warehouses to the cloud.

Course Outline

Cloud Data Warehousing – What and Why?

• The cloud migration trend
• Challenges of conventional data warehousing
• Overcoming the challenges

Benefits of Cloud Data Warehousing

• Scalability
• Elasticity
• Managed infrastructure
• Cost savings
• Processing speed
• Deployment speed
• Fault tolerance
• Disaster recovery
• Security and governance
• RDBMS in the cloud

Cloud Data Warehousing: Look Before You Leap

• Data movement
• Design and development
• Security and compliance
• Cloud and on-premises hybrid
• Portability

To Cloud or Not to Cloud

• Data warehouse architecture
• Technical debt
• Technology architecture
• Networking
• Data warehouse operations
• Data center operations
• Disaster recovery and business continuity
• Scalability and elasticity
• Data types
• Agility
• Governance
• Security
• Compliance
• Data quality
• Self-service
• Commitment

Cloud Migration SWOT Analysis

• Some guidelines
• Exercise 1 – SWOT for Migrating Your Data Warehouse
• Exercise 2 – Using the Migration SWOT Analysis
• Exercise 3 – SWOT for Not Migrating Your Data Warehouse
• Exercise 4 – Using the No Migration SWOT Analysis
• Exercise 5 – To Cloud or Not to Cloud: Making the Decision

Technologies for Cloud Data Warehousing

• Cloud data warehouse functions
  • Data storage
  • Database loading
  • Data processing
  • SQL queries
  • Workload management
  • Metadata management
  • Security and access controls
  • Data protection
• Cloud data warehouse platforms
• Migration tools
  • Integration Platform as a Service (iPaaS)
  • Data warehouse automation
  • Data virtualization

Step-by-Step Data Warehouse Migration

• The big picture
• The business case
• Architectural assessment
• Migration strategy
• Technology selection
• Migration
  • Planning
  • Schema migration
  • Data migration
  • ETL migration and data pipelines
  • Metadata migration
  • Migrating users and applications