Elizabeth Press

🚀 Exciting News! 🚀 We are thrilled to announce that Kumo AI's position paper "Relational Deep Learning: Graph Representation Learning on Relational Databases," has been accepted at #ICML 2024! 🎉 Join us in Vienna from July 21-27, where our very own Matthias Fey will be presenting this exciting new research area. 🔗 https://icml.cc/ In collaboration with Stanford University, we introduce Relational Deep Learning (#RDL), a new research area for applying deep learning on relational databases by viewing them as temporal, heterogeneous graphs. This method eliminates the need for manual feature engineering, leading to more accurate models built faster. 🚀 We’ve also developed #RelBench, a benchmark suite to facilitate research in this new area, using diverse datasets from Stack Exchange discussions to Amazon book reviews. 📚 Stay tuned for more updates on our presentation format, and check out our paper here: https://lnkd.in/dk24vbSg #ICML2024 #MachineLearning #AI #GraphNeuralNetworks #RelationalDeepLearning #DataScience #KumoAI #StanfordCollaboration

38

𝐒𝐩𝐞𝐧𝐝𝐚𝐭𝐚: A True 𝐄𝐧𝐭𝐞𝐫𝐩𝐫𝐢𝐬𝐞 𝐀𝐧𝐚𝐥𝐲𝐭𝐢𝐜𝐬 Solution (𝘐𝘵'𝘴 𝘯𝘰𝘵 𝘺𝘰𝘶𝘳 𝘨𝘳𝘢𝘯𝘥𝘱𝘢'𝘴 𝘴𝘱𝘦𝘯𝘥 𝘢𝘯𝘢𝘭𝘺𝘴𝘪𝘴 𝘢𝘯𝘺𝘮𝘰𝘳𝘦!) No "data warehouse", "data lake", or "data lakehouses" needed either! Nor do you need a star schema to avoid suffering in computational limbo either! (But if you still think PowerBI is a great presentation tool, it can build you a perfect star schema on millions of records in seconds ... ) Just real spend analysis, which includes the ability to: * extend the schema at any time * add new derived dimensions at any time * change those mappings at any time * build derivations, data views, and mappings that are dependent on other derivations, mappings, views, inputs, linked datasets, and so on, with real-time “recalculation” * create new views and reports WITHOUT dumping the data to Excel * preserve ALL of the above on cube data refreshes (that's right! you DON'T have to rebuild the cube from scratch on every refresh!) * do so in your OWN private copy of the cube (that inherits from the shared master cube) so you don't have to change the common cube for your own custom analysis and reporting needs * get the answer you need RIGHT NOW (not on the next refresh next month when you've forgotten why you were told to get the answer in the first place) So what else can it do besides every spend analysis report ever invented under the sun? * 𝐒𝐚𝐥𝐞𝐬 𝐃𝐢𝐬𝐜𝐨𝐮𝐧𝐭 𝐕𝐚𝐫𝐢𝐚𝐭𝐢𝐨𝐧 𝐀𝐧𝐚𝐥𝐲𝐬𝐢𝐬 Over Time By Salesperson … and Client Type ... that dynamically updates in real time on every parameter or "what-if" change * Real 𝐏𝐨𝐬𝐭-𝐌𝐞𝐫𝐠𝐞𝐫 𝐖𝐡𝐢𝐭𝐞-𝐒𝐩𝐚𝐜𝐞 𝐀𝐧𝐚𝐥𝐲𝐬𝐢𝐬 (if you think you can do this now, think again) * 𝐌𝐮𝐥𝐭𝐢-𝐯𝐚𝐫𝐢𝐚𝐭𝐞 𝐁𝐢𝐥𝐥 𝐑𝐚𝐭𝐞 𝐀𝐧𝐚𝐥𝐲𝐬𝐢𝐬 ... and more! https://lnkd.in/eM9Y4fez

9

📅 Mark your calendars for September 18th in #Vienna! At the #DataScience MeetUp, Andreas from Neo4j and Sebastian will talk about the relevance of #KnowledgeGraphs for Gen #AI, and explain why proper #GraphVisualization is critical for understanding your #data. We look forward to seeing you there! https://lnkd.in/e-EKZ9fZ

6

🔍 Just read a fascinating paper on the future of data ecosystem business models in the emergent "Artificial Intelligence of Thinks" (AIoT) space! Very useful systematic research that yields some key insights on how companies are transforming data into value in this developing ecosystem. https://lnkd.in/ert4nU7X 🔄 FOUR DISTINCT BUSINESS MODEL ARCHETYPES IN THE AGE OF AI ARISING 🤝 Adaptive Partnership - High control + High customization - Focuses on developing close, collaborative relationships with customers - Offers deep insights and personalized solutions - Best suited for strategic, long-term business relationships 👥 User-Centric - Low control + High customization - Prioritizes customer experience and flexibility - Enables broad market reach and scalability - Emphasizes user autonomy in configuring solutions ⚡ Standardized Efficiency - High control + Low customization - Focus on operational excellence and cost reduction - Streamlined processes and standardized offerings - Ideal for high-volume, routine operations [3] 🔄 Flexible Transaction - Low control + Low customization - Transaction-focused with minimal ongoing engagement - Emphasizes automation and self-service - Suitable for commoditized offerings 💎 AIOT BUSINESS MODEL VALUE CAPTURE STRATEGIES: 📊 Responsive Monitoring: Goes beyond traditional monitoring by enabling instant actions based on real-time data, including automatic corrective functions and predictive maintenance to prevent failures 🤖 Object Self-Service: Drives objects toward independent functionality with less human intervention, incorporating self-adjustment, replenishment, and self-operation capabilities 💡 Insights as a Service: A data-driven value proposition that leverages AIoT to collect, analyze and deliver real-time and collective insights that give meaning to massive amounts of data, enabling customers to make informed decisions   ✨ Enhanced User Experience: Focuses on improving customer interaction through digital add-ons and customizable interfaces and content that adapt to individual users' preferences and behaviors over time 💡💡 The research shows successful AIoT implementations balance control and customization to create sustainable competitive advantages. 💡💡 🤔 Where do you see your organization fitting in this framework? #AIoT #DigitalTransformation #Innovation #BusinessStrategy

15

1 Kommentar

🤖 Ever wondered how GPT "thinks"? Well... it doesn't. 😅 Imagine GPT as a text fortune teller with a bag of words. 🧙‍♂️✨ Every time you ask it something, it reaches into its magical bag 🛍️ and pulls out the next most likely word based on patterns it has "seen" before. 🔮 Ask it about space? 🌌 Out comes "galaxy," "astronaut," or maybe even "wormhole." 🛸 Ask it to count "r"s in a sentence? 🧐 It might give you a whole essay on why letters matter, but... does it really know how to count? 🤷‍♂️ GPT isn’t doing any deep thinking or reasoning, it’s just really good at playing the game of predict-the-next-word! 🎲 So next time it gives you a genius-sounding answer 🧠, remember: It's just the result of some serious text-juggling, not a brainwave! 😂 But hey, if juggling words was an Olympic sport, GPT would win gold 🥇 every time! Video Credits: 3Blue1Brown Great Learning Video for GPT on German follows in the comments. #AI #GPT #MachineLearning #LanguageLearningModels #DataManagement #Mindset #Leadership

18

5 Kommentare

Artificial intelligence continues to break boundaries, offering us unprecedented ways to explore and connect ideas across seemingly unrelated domains. A recent AI method developed by Markus J. Buehler at MIT exemplifies this potential. By integrating generative AI with graph-based computational tools, Buehler’s work bridges gaps between biology, art, and even music, revealing shared patterns of complexity and order. The AI model, grounded in concepts from category theory, can reason over scientific concepts symbolically. This allows it to make novel predictions and connections, such as identifying structural similarities between biological tissues and Beethoven’s "Symphony No. 9." The research, published in Machine Learning: Science and Technology, introduces a knowledge mapping system that turns large datasets, like collections of scientific papers, into highly connected graphs. These graphs enable the AI to suggest new materials, identify research gaps, and highlight previously unseen connections. In one compelling example, the AI suggested designing a new mycelium-based material inspired by abstract patterns in Wassily Kandinsky’s artwork. This material balances strength and adaptability, with applications ranging from sustainable construction to biomedical devices. This innovation underscores the immense potential of graph-based generative AI. It opens doors to interdisciplinary research that may drive breakthroughs in material science, bioengineering, and even philosophical inquiries about the intersection of art and technology. For more insights, visit the original report by MIT News.

1

🤠 Wrangling petabytes of unstructured data and metadata for #AI is a huge IT challenge! We're here to help. The backbone of every storage system is its #metadata database and efficiently managing #UnstructuredData is business-critical. 💡 The powerful integration of GRAU DATA GmbH MetadataHub + #ScalityRING empowers IT teams and developers to handle massive volumes of unstructured data, making it AI-ready. Transform how unstructured data and its associated metadata are managed to make it actionable with: ✳️ faster data access, ✳️ more visibility, ✳️ unlocked AI-ready workflows, ✳️ and improved AI outcomes. *************** Get the details in our blog: https://bit.ly/40t6JbI ************** P.S. Scaling metadata is one of the #10DimensionsOfScale that every storage architecture should be required to handle. You can learn more about #MultidimensionalScaling and #RedefiningScale for AI in the blog.

14

2 Kommentare

Building a Data-Informed Culture: Lessons from ION Analytics W. Edwards Deming once said, "It is not necessary to change. Survival is not mandatory.” I would argue that today, more than ever, a data-informed culture is essential for an organisation to survive. At ION Analytics, our journey toward building such a culture has been challenging, but our pursuit of data and insights over opinions has put us in a strong position to take advantage of the new wave of AI tools becoming available. Here are some key lessons we've learned along the way: [1] Set the Tone from the Top The foundation of a data-informed culture starts with strong leadership commitment. This means doing away with hunches, opinions, and seniority as the basis for decision-making, and letting the data help guide decisions. [2] Empower Every Employee Data democratisation is crucial for making data accessible to everyone in the organisation. Data is not something to be hoarded – it is to be shared. Centralise your data, make it accessible (and secure/permissionable), and get rid of the data gatekeepers. [3] Build Confidence and Competence Building a data-informed culture requires a workforce that is not only data-savvy but also confident in interpreting and leveraging data. From basic data skills to advanced analytics training, ensure every employee is equipped with the knowledge needed to harness the power of data. [4] Break Down Silos Collaboration between different teams is vital for maximising the value of data. By fostering a collaborative environment, you can ensure that diverse perspectives are considered, leading to more holistic and effective solutions. [5] Invest in Data Quality The adage "garbage in, garbage out" holds true in data science. Ensuring data quality is a non-negotiable aspect of any data strategy. Implement regular data audits, standardise data collection processes, and establish clear data ownership roles. [6] Question Everything Transitioning to a data-informed culture involves changing mindsets. Statements should only be made if they are supported by data; analysis plans should include an initial hypothesis that is sought to be validated; decisions should be made when backed by evidence. [7] Drink the Kool-Aid (and Eat Your Own Dog Food) Modern enterprises have an abundance of data, along with a range of ways in which improvements could be made by embracing AI. If you are going to claim to be an ‘AI-enabled’ organisation, leverage it internally before you try to sell it to clients. Building a data-informed culture is not a one-time project but an ongoing journey. The lessons we've learned underscore a fundamental truth: data is a powerful asset only when it is leveraged effectively – and your ability to take full advantage of the benefits are greatly increased when everyone in the organisation is part of the journey. #AI

26

2 Kommentare

Making Sense of Data Context: A Simple Guide We often see LinkedIn posts and read about "understanding data context" but what does this really mean in practice? For me, it boils down to asking these three questions in a respectful, safe environment where no one feels offended: • Inquire about data sources: • Would you mind sharing the sources of this data? • Once you know the sources, ask: • Who can tell me which systems touched this data along along the way? • Before delivering the dashboards or reports ask: • Why does it matter to our users? Pro tip: Ask these questions early in your data projects. They'll save you time, build trust with stakeholders, and ensure your analysis actually drives value. Remember: Better context leads to better decisions but make sure the environment is safe before you shoot your questions. What other things you do to learn the data context better? Please share. #businessIntelligence #dataContext #databestpractices #dataleadership #bi #dataengineering

6

At Frankfurt: Programming Related to Artificial Intelligence - Publishing Perspectives: At Frankfurt: Programming Related to Artificial Intelligence  Publishing Perspectives http://dlvr.it/TC141m #ai #artificialintelligence

1

#EarlyCareer #talk at #IJCAI2024 by🗣️Julia Neidhart, Technische Universität Wien: Transforming Recommender Systems: Balancing Personalization, Fairness, and Human Values ➡️https://lnkd.in/dK3Nieyr Abstract: Recent advancements in recommender systems highlight metrics beyond accuracy, including diversity, serendipity and fairness. This paper examines the integration of Large Language Models into recommender systems, focusing on challenges such as interpretability, bias, and user autonomy. LLMs offer enhanced personalization capabilities but also introduce concerns regarding transparency and ethical implications. This work discusses ongoing research efforts aiming to develop transparent, fair, and contextually aware systems. Our approach seeks to prioritize user wellbeing and inclusivity, contributing to a more equitable and functional digital environment through advanced technologies and interdisciplinary insights. #XAI #LLMs #InclusiveTech #NextGenAI #AI TU Wien Informatics

3

Welcome to read the article "Integrated Gradients for Feature Assessment in Point Cloud-Based Data Sets" which is written by Mr. Markus Schwegler, Prof. Christoph Müller and Alexander Reiterer. This excellent article was published in issue 7 in 2023 and has been viewed 1313 times and cited 3 times! Article link: https://lnkd.in/gMCsed5z MDPI Fraunhofer IPM Furtwangen University Albert-Ludwigs-Universität Freiburg (Albert Ludwig University of Freiburg) #pointcloud #neuralnetwork #integratedgradients

2

Is a Data Modeller Really Necessary? Absolutely! Here’s Why: In the age of big data and digital transformation, the role of a Data Modeller has become more critical than ever. But why exactly do we need Data Modellers? Key Reasons Why Data Modellers Are Essential: 🔍 Organizing Complexity: As organizations collect vast amounts of data from various sources, Data Modellers create structured frameworks that simplify this complexity. They turn chaotic data into a well-organized asset. 📊 Informed Decision-Making: A well-designed data model allows businesses to extract actionable insights quickly and accurately. This drives better decision-making across all levels of the organization. ⚙️ Optimizing Performance: Efficient data models enhance database performance, ensuring that data retrieval is fast and reliable, which is crucial for real-time analytics. 🔒 Maintaining Data Integrity: Data Modellers set standards and rules that ensure data consistency and integrity, preventing errors and maintaining high-quality data across the organization. 🌐 Future-Proofing: As businesses grow and technology evolves, scalable and adaptable data models are essential. Data Modellers design systems that can handle future data needs without major overhauls. 🤝 Bridging Gaps: Data Modellers act as a bridge between technical and business teams. They translate business requirements into technical specifications, ensuring that data systems meet business goals. In essence, Data Modellers are indispensable in harnessing the full potential of data. They lay the groundwork that allows data scientists, analysts, and business leaders to do their best work. Without their expertise, the path from raw data to strategic insight would be far less efficient and far more error-prone. Let’s Appreciate Our Data Modellers: If you work with data or rely on data-driven decisions, take a moment to appreciate the Data Modellers who make it all possible. Their work might be behind the scenes, but its impact is front and centre. Question for You: How has effective data modelling made a difference in your organization or project? Share your experiences and let's discuss the impact of this crucial role! 📊🔍 #DataModelling #DataScience #DataIntegrity #DataDriven #TechHeroes #DigitalTransformation

39

2 Kommentare

The Rails 8 feature set perfectly complements SQLite's power in creating resilient, high-performance production apps, but still the question lingers: Can I really go all-in on #SQLite? Join Stephen Margheim at #RailsWorld as he illustrates how to leverage Rails and SQLite's full potential in your next venture, and when SQLite does and doesn't make sense for your application. https://lnkd.in/eaBstScy

52

1 Kommentar

Following suits of great people in my network and seeing a trend almost before we bit goodbye to 2024, so here ny my 2024 LinkedIn Rewind, by Coauthor.studio: 2024: The year I discovered that building world-class data platforms isn't about perfect technology—it's about creating an environment where human imperfection becomes our greatest innovation engine. This journey transformed how I see leadership in AI and data. No longer just about technical excellence, but about understanding how diverse perspectives and unique human strengths can solve complex challenges. My transition to Director of AI and Data Platform at Collinson wasn't just a title change; it was a profound shift in understanding that breakthrough innovations emerge from teams who embrace their differences. Three posts this year crystallized this transformation: ➡️ Product Development Insight: "An underappreciated engineer and an overwhelmed product manager—two sides of the same coin, yet often miles apart." Success comes from bridging technical excellence with human understanding. https://lnkd.in/efJE8Yub ➡️ Team Building Reflection: "Building a great team means bringing together great people. Not everyone needs to fit perfectly into your ideal checklist; what someone lacks in one area often highlights their strengths in another." https://lnkd.in/egEV6qf4 ➡️ Leadership Philosophy: "Make it a habit to ask at least one person every day, 'How can I help?'" https://lnkd.in/ekWHy7ap Beyond these insights, 2024 marked significant milestones: completing the INSEAD Leadership Communication with Impact certification, speaking at the CDO Travel and Hospitality Exchange, and continuing to grow "One Perspective at Once" on Substack. Looking ahead to 2025, my focus remains unwavering: building platforms that unifies the intelligence both of human and machines, that don't just process information, but unlock true potential. The most sophisticated AI won't replace human creativity—it will amplify our collective intelligence. For data leaders and AI practitioners: your most powerful technology is the team you build, not just the code you write. #AI #DataLeadership #LeadershipLessons #TeamBuilding #LinkedInRewind #Coauthor #2024wrapped

19

Multivariate Time Series (MTS) Forecasting: A Persistent Challenge in Receivables Management - Could the Dozerformer Be the Breakthrough? Accurate solutions for predicting bad debts and general debtor behavior are often proclaimed, but the reality is that it remains an unsolved problem. This challenge is not only a fascinating scientific pursuit but also the holy grail in the field of predictive analytics for receivables management. A truly effective method for accurately predicting bad debts or recovery rates would instantly have a massive impact on cost efficiency and debt recovery success. However, this issue proves to be extremely persistent. Predicting default rates and probabilities during the collection process deals with high-dimensional and strongly correlated data, and the temporal dimension must be considered at the level of events, debtor data, and payment behavior. As someone who has been dealing with this topic for many years, I have tried a multitude of methodologies myself, including classical attention or LSTM networks. While some of these methods have led to significant successes and substantial business impact, it has been partial at best. Therefore, it is all the more exciting that a new attention framework, the so-called Dozerformer, has been developed. This novel framework is designed to tackle the complex problem of Multivariate Time Series (MTS) forecasting. The proposed Dozer Attention mechanism, a key feature of the framework, is made up of three components: Local, Stride, and Vary. These components are designed to capture the essential attributes of MTS data, such as locality, seasonality, and global temporal dependencies, and adaptively utilize historical time steps based on the forecasting horizon. The Dozerformer was evaluated on nine benchmark datasets and demonstrated superior performance in both accuracy and efficiency compared to recent state-of-the-art methods. The study also includes an ablation study to validate the effectiveness of the Dozer Attention mechanism. The research concludes that the Dozerformer significantly outperforms current methods, making it a promising tool for solving complex and real problems in modeling and predicting multivariate time series. This development is particularly intriguing, and I am eager to see if this novel approach will truly facilitate a breakthrough for the transformer architecture in addressing complex, real-world challenges in creating models and forecasts for multivariate time series, particularly in the context of financial services and receivables management. #AI #genAI #generativeAI #artificialintelligence #PredictiveAnalytics #TimeSeries #Forecasting #TechNews #DebtCollection #creditrisk #CreditManagement #ClaimsManagement #quantitativefinance  #financialservices https://lnkd.in/eWRjpYWk

12

1 Kommentar

One feature I have been looking for in #Quarto has been support for Scrollytelling. At posit::conf(2024), Andrew Bray has just presented #Closeread - a Quarto extension that enables quite a few scrollytelling techniques in Quarto. Closeread comes with different layouts and support for cross-fading plots, zooming in on code and images, and highlighted parts of a text. 🤯 Quarto is an open-source publishing system developed by Posit PBC that allows users to create and share dynamic, reproducible content like articles, presentations, dashboards, and books using #Python, #rstats, #Julia, and more, in various formats including HTML, PDF, and Word. The screengrab shows the example of Edward Tufte's description of Charles Minard's famous visualization of Napoleon's disastrous military campaign in Russia from the closeread documentation. I can't wait to try this out! #PositConf2024 #DataScience #dataviz

435

42 Kommentare

Discover potential paradigm changes for data and analytics💡 BISE Journal recently published a new article by Ivo Blohm, Felix Wortmann, Christine Legner, and our own Dr. Felix Köbler. 🔎Check out the article "Data products, data mesh, and data fabric" which explores these concepts, highlighting their potential to revolutionize data management and analytics: https://lnkd.in/e7F-z6p6 #DataProducts #DataMesh #DataFabric #DataProductManagement

11