The importance of facing, understanding, predicting and even mitigating uncertainty has been well acknowledged and studied in various fields such as philosophy, physics and finance. In terms of software and system engineering, due to the increasing complex of large-scale systems themselves, and the dynamic and unpredictable deployment and operation environments of such systems, increasing attentions have been given to address challenges of explicitly specifying and modeling uncertainty.
Uncertainty Modeling (UM) aims to promote and enable explicit specifications of uncertainty and uncertainty related concepts, at various contexts (e.g., developing large-scale Cyber-Physical Systems and Internet of Things, for different purposes (e.g., enabling uncertainty-wise requirements specifications, modeling, verification and validation, e.g., facilitating the definition of uncertainty-wise testing strategies), and at different phases of the development of such complex and uncertainty-inherent systems (e.g., requirements, architecture, design, testing and operation).
In the context of OMG, we see a diverse set of uncertainty modeling applications, such as 1) integrating with UML use cases, SysML Requirements Diagram, to enable requirements V&V, 2) capturing uncertainty as part of SysML or UML models to facilitate design and/or testing, 3) integrating with BPMN and other OMG standards to facilitate different kinds of analyses and generations.
OMG’s Uncertainty Modeling Request for Information (RFI) is currently open for responses. The RFI aims to solicit ideas, discussions, comments, recommendations, user needs and experiences about uncertainty modeling. Collected responses will be carefully analyzed and will be used to identify requirements, based on an RFP for an UM will be developed. Instructions for responding to this RFI are specified in the OMG Uncertainty Modeling Request for Information document (ad/16-09-02 (Uncertainty RFI)).
We invite you to join the conversation.
Presented by one of the foremost experts in BPM standards, this session will concretely demonstrate usage of the three leading business modeling standards produced by the Object Management Group (OMG). This session will explore the positioning and core behavioral differences between the Business Process Model and Notation (BPMN), the Case Management Model and Notation (CMMN) and the Decision Model Notation (DMN). The specific roles and usage of these dominant business modeling notations will be explained and demonstrated using a worked out example integrating BPMN, CMMN and DMN models.
What exactly are BPMN, CMMN and DMN?
When is one of these standards best suited for the purpose?
How to use BPMN, CMMN and DMN together?
What are the best practices for these standards?
To succeed, an analytics or data science team must effectively engage with business experts who are often inexperienced with advanced analytics, machine learning and data science. They need a framework for connecting business problems to possible analytics solutions. Decision modeling brings clarity to analytics projects, linking analytics solutions to business problems to deliver value.
In this webinar, IIA Expert James Taylor shares four key lessons learned by the central analytics team at a global leader in information technology. These lessons underscore that decision modeling builds a shared understanding with their business clients, revives projects that have lost their purpose, brings clarity to problems long thought difficult and delivers value quickly. The webinar will also introduce the basics of decision modeling and provide practical recommendations for adoption. A Leading Practice Brief based on this case study will be available for all registrants.
Ensuring analytics projects have clarity of purpose is critical to delivering business value. Learn how decision modeling helps a leading organization bring focus to its data science and analytic investments.
The focus of modern business intelligence has been self-service; pushing data into the hands of end users more quickly with more accessible user interfaces so they can get answers fast and on their own. This has helped alleviate a major BI pain point: centralized, IT-dominated solutions have been too slow and too brittle to serve the business.
What has been masked is a lack of innovation in data modeling. Data modeling is a huge, valuable component of BI that has been largely neglected. In this webinar, we discuss Looker’s novel approach to data modeling and how it powers a data exploration environment with unprecedented depth and agility.
Topics covered include:
-A new architecture beyond direct connect
-Language-based, git-integrated data modeling
-Abstractions that make SQL more powerful and more efficient
Data Modeling is a critical component of enterprise BI. Most data modeling is desktop-based. This can lead to a large number of problems including scale, maintenance, collaboration and security.
Join Chris Webb independent consultant and Peter Sprague VP Solutions Engineering at Pyramid Analytics for a discussion about how to avoid an epidemic of poorly designed and scattered data models throughout an organization. Questions tackled will include:
-What is the impact of poor data modeling decisions?
-Can data modeling decisions be delayed?
-Do new tools and technologies alleviate the need for good data modeling?
-What are the data modeling needs of an enterprise?
Boost your modeling performance. This simple demonstration of ensemble modeling will show you one way how. Intended for marketers and all levels of analyst.Read more >
It’s probably not too often that you’ll get this perspective. Star Wars was really all about information disclosure threats! You’ll want to find out more as noted presenter and author Adam Shostack, references one of George Lucas’ epic sagas to deliver lessons on threat modeling. Not only was the Death Star badly threat modeled, the politics between Darth Vader and Gran Moff Tarkin distracted incidence response after the plans were stolen. This session will provide you with proven foundations for effective threat modeling as you develop and deploy systems. Adam will help you understand what works for threat modeling and how various approaches conflict or align. The force is strong with this session.Read more >
Marketers have been effectively analyzing their own customer data and identifying similar buyer attributes offline for decades. Unfortunately customer data is often siloed or spread across dozens of applications, preventing marketers from leveraging insights that should be used to model the perfect new audience for digital campaigns, or to campaign performance against offline consumer sales.
When your data is connected insights can be shared across marketing applications and measured offline or online. Data from every marketing technology in your toolbox can finally be connected allowing you to continuously refine your campaigns. Sign up to join our next webinar where you’ll learn:
– About new strategies for audience modeling and practical tips for getting started
– Reach the perfect audience by connecting your customer data to digital marketing campaigns
– To forget about measuring campaign performance with vanity metrics like “reach” or “impressions”
– How to determine which digital campaigns are actually delivering measurable sales
Understanding data modeling can help you get the best insights out of your data. The challenge of data modeling is to understand how to work with complex data in order to standardize, structure and optimize data to gets accurate insights quickly.Read more >
The USP of Hadoop over traditional RDBMS is "Schema on Read".
While the flexibility of choices in data organization, storage, compression and formats in Hadoop makes it easy to process data, understanding the impact of these choices on search, performance and usability allows better design patterns.
Learning when and how to use schemas and data model evolution due to required changes is key to building data-driven applications.
This webinar will explore the various options available and their impact to allow better design choices for data processing and metadata management in Hadoop.
BI enables human decisions based on facts, but Predictive Modeling enables computer decisions based on probability. From our spam folders to our Netflix recommendations, to our credit scores, we are all touched by predictive decision automation every day without realizing it.
But now, thanks to shrinking compute costs and growing data availability, predictive models can even automate business decisions in Marketing, Risk, Fraud, Security and Quality, dramatically increasing both speed and accuracy. In this thought-provoking talk, Tim Negris, from machine learning software pioneer Yottamine Analytics, will deconstruct and demystify predictive modeling and show how nearly any business can use it to optimize profit, responsiveness, and efficiency.
Learn to Store and Query Times Series Data in NoSQL and Other Use CasesRead more >
Discover how businesses turn big data into meaningful insights to help make organizations work smarter, and make better decisions faster.
Join Scott Dallon to learn tips on analyzing and modeling complex data sets!
Understand profitability to provide profit scenarios and identify leading indicators for success.Read more >