Review Agenda Machine Learning Week Europe 2022

Berlin - October 5-6, 2022

Transforming data into value is a strategic focus for companies across industries. With the gap between leading data driven businesses and late movers growing rapidly, decision makers are wondering what secret ingredient helps companies to successfully leapfrog roadblocks on their value journey. Since this is evidently not a one-dimensional challenge, we’ll take a hands-on perspective on what really helped companies to sustainably turn data into value.

International FMCG manufacturer Beiersdorf needs to forecast 1000s of products every month. In 2021, 10 years after the 1st Neural Networks were introduced, Beiersdorf set out to improve automatic forecasting further by reviewing the latest developments in technology. Surprisingly, some of the most recent and hyped algorithms such as DeepLearning, XGBoost, Prophet, BSTS and others did not perform well, but simple AI-methods customised to their data properties improved accuracy significantly.

This keynote gives an overview of the current innovations within the realm of Natural Language Processing. It will cover the various applications which can be built using the Transformer architecture and dives into the importance of context for AI in today’s world. This will be followed by an in-depth overview of Semantic Search and how to build it from scratch and develop an API for it.

Decision-making is about leveraging data & AI to manage volatility in the crazy world we  live in. Forward-looking decisions leveraging AI-based forecasts are oftentimes seen as a “black box”, which slows down utilization. Prof. Dr. Frauke Schleer-van Gellecom will demonstrate how the most complex trade off between accuracy and explainability can be handled and how the black box can be transformed into a grey box.

Many manufacturing companies have piloted predictive analytics solutions for Industry 4.0 use cases, but only few have yet managed to industrialize them at scale across a large production network. Michael Bruns and Andreas Odenkirchen will take a look at the common pitfalls and share their best practices from working with a global automotive supplier to deploy machine learning models at scale using a digital manufacturing platform foundation that leverages Cloud, IoT, MLOps and DevOps capabilities.

A newspaper subscription is defined by various critical events, ranging from the end of the trial subscription to receiving invoices. Based on predictive analyses that anticipate customer behavior during these events, we develop, test, and implement customized marketing interventions covering the whole subscription lifecycle. You will learn about modeling via a custom AutoML-pipeline and its close intertwining with marketing execution that aim to maximize subscription lifetime value at DIE ZEIT.

A malt yield forecast with an excellent prediction performance, as well as first transfers to Augustiner Bräu, were successfully implemented to optimize the beer brewing process. The crucial next step is getting our ready-to-use analysis modules with built-in requirements into the running production. For this, we are creating an architecture for robust deployment, considering model resilience and automated detection of data drift and performance decay to eventually trigger new model training.

Despite the wealth of information available to job seekers, choosing one’s next job remains a complex endeavor. We will present an approach for creating job recommendations across industries by combining hand-annotated job data and a multivariate technique combining job title, job description, and skill set similarity. This method was used to scan 9 million job transitions to identify the best matches for any given job and uncover deep insights into how the labor market is organized today.

Global enterprises are increasingly relying on data analytics for decision making. To process data, firms leverage cloud-based data warehouses. As more on-prem data is moved to the cloud, the need for robust data governance controls to ensure data integrity, security, and regulatory adherence is mounting; however, existing governance processes are lagging. Here we present a zero-trust approach that can augment existing governance models and reduce exposure of sensitive data like PII.

Survival analysis, the modeling of time-to-event data, is a statistical field with a long history and great potential to marketing and analytics. In this deep dive, you will learn about the brief origins of survival analysis and applications in the field of customer retention, which is of particular importance for subscription-based growth. Learn how to grow your understanding of customer churn, and learn how to better predict your customer’s lifetime, along with monetary aspects altogether.

The sound-based machine learning solution developed for Hindustan Shipping Limited helps identify defects that are sub surface or interior to the part. The identification of defect is real time, during the part production. Thus, enabling one to take actions immediately and not wait to produce a scrap part. Takeaways of this presenation are:(1) insights into the sound based machine learning approach for sub surface and interior defect detection; (2) how to identify the location and magnitude of defect in real time.

As a manufacturer, visual inspection is a crucial part of keeping the quality of our products up to high levels. Above that many new applications can be done by making use of deep learning in combination with computer vision. Deep learning models can be industrialized with the Visual Perception Provider (VPP) which is a service developed in-house at Continental Tires. The talk is about the reasoning, why we decided to go that route, and about what we are doing. Also, some use cases done with the Visual Perception Provider will be demonstrated.

In recent years, there have been various efforts to product thinking and decentralized data loading using data meshes. However in deep learning, data loading is still challenging to master at scale. In this talk, we present our decentralized data loading solution and show why flexibility and collaboration is key to enable novel ML use cases. We hope to make large-scale model training accessible to a wider community and move towards more sustainable ML.

To evaluate the goodness of models we have a whole lot of KPI with a set of underlying ideas of “what goodness is”. However, the impact on the real-world (when the model is applied there) is very rarely taken systematically into account when the goodness of models is evaluated. Here we show a simple idea to assess the expected improvement (win) an a real-word situation for any classification problem at the example of campaign optimization and credit rating. Of course, this also allows to compare the real-world impact on win of different classifiers.

By predicting problems the laser machine availability can be increased significantly. This will not only reduce the costs of the maintenance. Started as a pure condition monitoring portal, the project for Trumpf Laser evolved into a hollistic predictive maintenance system, which allows facilitating the work of other departments (e.g. customer support).It also was the starting point for a new service: proactive support. Those practical examples shows the importance of empowering data-driven intelligence for machine manufacturers.

This session presents Anomaly Detection Neuroevolution (AD-NEv) – a multi-level optimized neuroevolution framework. The method adapts genetic algorithms for: i) creating an ensemble model based on the bagging technique; ii) optimizing the topology of single anomaly detection models; iii) non-gradient fine-tuning network parameters. The results prove that the models created by AD-NEv achieve significantly better results than the well-known anomaly detection deep learning models.

During this session, we will discuss the different risk-based categories of AI laid out by the EU’s Artificial Intelligence Act and find out how to become more admissible as per the Act. Thereafter, we will walk through the concrete steps, tools, and practices such as monitoring, explainability, model fairness, and compliance that are instrumental in achieving Responsible AI and building more risk-proof and market-friendly solutions

Geo-experiments – advertising experiments where the treatment and control groups are chosen based on users’ locations – provide a privacy-friendly alternative to cookie-based online experiments that can also be used to measure offline effects of online advertising. This session discusses the algorithms we use at Google to design the experiment regions based on geographical user behavior, and the rigorous statistical methods to analyze randomized experiments based on these regions.

Estimating lifetime of a machine or a wearing-part in a complex machine is a requirement in cost-optimal maintenance planning to reduce costly downtime or avoid too frequent maintenance. Regression models mapping features to the time-to-failure are not suited in the real world because of censored data. This talk shows how survival regressions can be utilized for a maintenance planning application. We discuss and demonstrate available survival analysis tools, their strengths and limitations.

Being able to predict KPIs for new, not yet published articles is a key factor in process optimization to assist editors with their daily work flow. Modern NLP methods allow us to use text content efficiently and to understand the connection between natural language and the respective KPIs. By combining statistical models (GAM), modern NLP methods (BERT), and XAI tools (SHAP), we are able to understand specific connections between text content and our KPI.

Machine learning models are becoming obsolete and must be retrained – this is the current widespread tenor. Is this actually true? And if yes, which components does a CI/CD pipeline for machine learning really need – and which are optional? How can the whole thing be implemented without building a complete Machine Learning Platform team? And which challenges are still difficult to solve at present? A field report including (mis)decisions, which will help to choose the right path for your own challenges.

This talk will introduce the foundational business skills you’ll need to deliver business value and grow your career as an analyst. Drawing on best practices, published research, case studies and personal anecdotes from two decades of industry experience, we give an overview of foundational skills related to Company, Colleagues, Storytelling, Expectations, Results and Careers–emphasizing how each topic relates to your unique position as an analytics professional within a larger corporation.

The desire to embody Responsible AI practices requires an understanding of, the context around, and the impact on the end user. To achieve this, design and research are just as pivotal to the RAI conversation as ML. There is no bigger risk, and no greater irresponsibility, than to not interface with those who will be affected by your design. This talk will share how to navigate customer relationships to encourage end user contact and mitigate assumptions and therefore risk.

In January 2020, just after the first case of COVID-19 was discovered in the US, a multidisciplinary team of AI and medicine experts both in the US and China developed the first COVID-19 Clinical Severity Predictive Analytics Tool. In July 2021, another tool named COVID-19 Early-alerts Signals was built on a digital epidemiology framework that analyses alternative data sources to discover predictors of the pandemic curve. After the vaccine rollout, another team also co-led by the speaker developed a Vaccine Hesitancy Analytics Tool which is a real-time big data analytics cloud application to track misinformation and extract themes and topics related to vaccine hesitancy. In this keynote, Prof. Bari will outline the experimental research results from the three studies and the tools his team developed at New York University. This talk will also cover how governments and health systems can detect and respond to outbreaks, and how they can prepare for future pandemics using predictive analytics, mature AI capabilities, and infectious disease knowledge.

Relying solely on coding data scientists to handle your company’s predictive analytics projects is a recipe for failure. No company in the
world can hire enough data scientists to support all their use cases, and leaving domain experts out of the process doesn’t just overload your analytics teams—it also ignores critical business context. 

Join RapidMiner Account Executive Friedrich Rath to learn: 

·        How your organization can leverage the right technology and approach to get non-coding domain experts contributing to AI projects 

·        Why upskilling your people is the key to executing faster and building truly differentiated data-driven solutions 

·        How data scientists and domain experts can work together to bring more models into production throughout the business and scale the impact of AI enterprise-wide 

RapidMiner was started by PhD data scientists who understood that the power of AI shouldn’t be reserved for…well, PhD data scientists. RapidMiner is a code-optional data science platform that can support
anyone in your organization across the full analytics lifecycle—everything from making sense of your data to building a model with it to ultimately deploying an AI-powered app to drive better decision-making. 

In this session, Dr. Yvonne Blum will present how machine learning techniques in MATLAB were used to detect error conditions in MAN Energy Solutions gas turbines. These engines are distributed all over the world, quite often located in very remote areas where machine failure can have severe consequences. The goal of the project was to automate the time-consuming process of visualizing and manually evaluating measured sensor data, to determine gas turbine error conditions at an early stage. 

The case study was authored Dr. Holger Huitenga and Dr. Yvonne Blum.

Immobilienscout24 went from outsourcing Marketing PPC activities to creating their own Marketing and Data capabilities. The data lake enables cross channel performance dashboards, predict sales and spending volume, create automated campaign alerts and outlier detection. In this case study, we will outline the lessons learned and the impact it had on both marketing and data teams.

Assembling a car is a complex task consisting of many steps usually grouped and organized in work plans. Based on the car model and its specifications, creating a key performance indicator (KPI) optimized work plan can be very time consuming. This case study at Volkswagen shows how data science can assist and speed up this process. After using various text analytics methods to identify similar work plans descriptions, a semi-automated benchmarking approach provides a KPI-driven recommendation.

In my quest to become a Pokémon master, I need to learn a lot about their types. I’d rather create a model to do so for me. A simple one-off won’t do: to be ready for upcoming generations, I need a pipeline for experimenting with new datasets and configs. We will set up a codified ML pipeline using the open-source DVC library. This will help us adopt a reproducible, experiment-driven approach to ML, which will boost our ability to iterate over models and compare them to find the best one.

Over the past two years, ONE LOGIC supported a major Covid vaccine manufacturer and several government agencies in using data to reliably scale the Covid vaccine production and distribute doses to when and where they are needed. We believe this case study can be applied broadly across sensitive supply chains in biotech and pharma, but also on a global scale for issues like the current “chip crisis”.

AI technology allows improving the UX significantly by showing the most relevant personalized content in real-time depending on the user’s browsing activity. By using Matrix Factorization with the Implicit Alternating Least Squares method we can calculate individual ratings for every user. Those ratings can be used to rank all the contents available on the website, and only the most relevant content is shown to each single user. The result: increased engagement and a lead rate uplift of 42%.

Quality improvement for mass production lines is an ongoing topic for many years. The target is to reduce the rate of defects and scrap during production having an impact on sustainability, delivery time and cost. For the example of mass lens production at ZEISS we introduce a Markov-based method, that allows us to trace the movement of a given product through the production line to help us understand potential root causes for quality losses and thus being able to predict defects. In the end we aim to achieve a closed loop quality control avoiding quality losses by an improved understanding of root causes and proactive actions enabled by Industry 4.0 technologies such as Machine Connectivity and Artificial Intelligence.

In this session, two open source libraries will be demonstrated to show you how you can quickly deploy custom Named Entity Recognition (NER) solutions. NERDA is an easy-to-use interface to apply pre-trained transformer models (e.g. based on Huggingface) to your own challenges. Combined with FastAPI, a lightweight webframework for Python, you can build your solution in short time

Sensing customer demand accurately is a crucial factor to optimize the inventory levels in the retail sector. Using sophisticated black box machine learning algorithms provides our client, a major Turkish retailer, accurate forecasts. However, it makes it impossible to understand the effect of each feature for model outputs. Our solution for Explainable AI (xAI) solved this problem by providing transparency on the feature level and enabled the retailer team to trigger actions for managing the customer demand in the market effectively.

After quick successful POCs, the productive rollout of AI solutions often comes with unexpected challenges, especially for human-in-the-loop applications. The presentation will illustrate a holistic solution in three sections, starting with an end-to-end overview with the example of ML-based assistance systems, followed by deep-dives into the two main pain points: Imputation approaches for dealing with imperfect data and the psychology behind motivators for human interaction with such systems.

This session is an introduction into Apache TVM – an end to end compiler framework for deep learning models. It can compile machine learning models from various deep learning frameworks to machine code for different type of hardware targets like CPU, GPU, FPGAs, microcontrollers. It provides bindings for different higher level languages like C++, Rust etc. and also has provision for autotuning the models for different hardware targets. You will learn how to use Apache TVM to deploy your models on different target systems.

This table discussion focuses on how data scientists, data nerds, head of data science, and chief data officers can contribute to sustainability in AI. We would like to discuss this from different views like science contribution to SDGs as well as running climate-neutral model developments. The participants should get some understanding on the carbon footprint of doing intensive modeling and what the industry can do to reduce it.

Unicode has unified characters from world languages, symbols and emoji into a single standard enabling easy interoperable communications. However, in recent years the availability of similar symbols from diverse languages is being exploited to deceive users for malicious ends. Deep learning has given us the tool to prevent these attacks. You will be shown how this exploit works and how to detect and prevent it by training a Deep learning model to detect visual similarities between characters.

Healthcare is about people- the patients receiving care, the people delivering it and those creating ways to support this. Data via predictive analytics is an enabler to supports a healthcare provider or system to achieve operational excellence to help problems thus faced by the people. Doing this well not only delivers patient outcomes but supports preventative and proactive population health management. In this table discussion, we would like to discuss these questions: What is the connection between operational excellence, predictive data analytics and artificial intelligence? What are the driving forces for achieving organizational performance by using artificial intelligence? What are the barriers to achieving organizational performance using artificial intelligence?

The latest formula for making sound investment decisions involves mining new alternative data sources, using predictive analytics, swarm intelligence, reinforcement learning, and high-performance computing. In this talk, Prof. Anasse Bari explains how those components are driving value in the world of finance and how new Artificial Intelligence algorithms are reinventing Wall Street. He will filter fact from fiction, and outline successful use cases that he has recently led (e.g. how social performance and consumer reviews could be used as predictive features, how to derive actionable insights from geospatial images.). Prof. Bari will also present an overview that can help you design an AI strategy and implement viable solutions to generate a “predictive analytics-based investment thesis.”

Telefónica, the biggest Spanish telecommunications company, asked us to provide a machine learning solution capable of detecting when one of their DSLAMs has an anomaly in registered customers indicating a loss in customer IPTV service. You will be shown how to deal with thousands of time series data by combining clustering algorithms, smoothing methods and deep learning tools to obtain efficient and high-performance results.

Many safety-critical systems, such as autonomous vehicles, rely on neural networks as state-of-the-art for image classification. Despite high accuracy, their final classification decision is difficult to verify. Uncertainty on how neural networks will behave is a challenge to safety. A known issue is providing high probabilities for unknown images. This deep dive will present some novel solutions to inspect certain parameters of trained neural network for detecting out-of-distribution data.

While pharma companies are increasingly realizing the value of data, they need to realize the importance of data mindset among non-data employees. Without the right mindset neither BI nor AI will be utilized. MSD has implemented a 3-year data literacy program to enable marketing and sales employees to understand customer data & insights and know how to use it. We are in the middle of this journey and will share the approaches we have used and our experience so far.