Take advantage of these pre-conference exhibitor workshops on for a hands-on demonstration of the latest industry products and software. All attendees are welcome to join onsite or pre-register. If you’ve already registered for the Annual Meeting, you can add workshops by editing your record. All Exhibitor Workshops will be held in the Phoenix Convention Center on Saturday, October 14.
Location: CC-North 121 A
Multi-Sigma: an Integrated AI Platform of Prediction and Optimization for Multiple Target Objectives Simultaneously
Presented by: Kotaro Kawajiri
AIZOTH provides AI services such as Multi-Sigma, AI consulting, spot support to optimize manufacturing conditions, and commissioned R&D. Multi-Sigma is the cloud-based AI software for R&D to reduce the effort of experiment drastically and also to help researchers finding the innovative solutions for their actual problems with minimum experimental dataset. Multi-Sigma was already introduced by large manufacturing enterprises and top universities. We will demonstrate how a Multi-Sigma can be used with sample datasets.
Key features of Multi-Sigma:
- Numerical analysis using deep learning: Deep learning techniques of both neural network and Bayesian optimization can be used with minimum dataset. High precision prediction of multiple objective variables. Factor analysis using sensitivity analysis for explainable AI. Multi-objective optimization in trade-off (maximization, minimization, target value, tailor-made optimization, constraint of explanatory variables). It can handle a large number of explanatory variables (up to 200) and objective variables (up to 100),
- No-code cloud-based software: Multi-Sigma is the advanced cloud-based AI platform which can be used with no-code on the browser. Anyone can use Multi-Sigma anytime, anywhere, and by any hardware (even by smartphones).
- Innovative Design of Experiment using Artificial Intelligence (AI-DOE): Researchers can utilize our ‘AI-DOE’ techniques easily on Multi-Sigma. AI-DOE is developed by incorporating AI techniques into the traditional DOE basing on statistics. It can guides researchers to find the optimal solution for multi-input and multi-objective systems.
- Versatility: Multi-Sigma can be utilized for any issue of R&D. Main users are manufacturing companies such as automobile, chemical, pharmaceutical companies. Also it can be used for management issues such as prediction of sales, marketing, and inventory management. Tailor-made optimization is the brand-new techniques with great potential in the field of medical care, material science, and sales management.
Location: CC-North 121 A
Teaching, Learning, and Applying Optimization: New Developments in the AMPL Modeling System
Presented by: Filipe Brandão and Robert Fourer
Optimization is the most widely adopted technology of Operations Research and Analytics, yet it must steadily evolve to remain relevant. After an introductory example, this presentation takes you on a tour through new developments in the AMPL modeling language and system that have been changing the ways that people learn and apply large-scale optimization:
- A new approach to connecting the modeling language to solvers, which lets you write many common logical conditions and “not quite linear” functions in a much more natural way, avoiding complicated and error-prone reformulations.
- A Python-first alternative to learning AMPL and model-building, supported by new teaching materials that leverage the power of Jupyter notebooks and Google Colab to bring modern computing to the study of optimization.
- Enhancements to the AMPL Python interface (amplpy), including faster data input and closer solver integration, which expand the possibilities for model-based application development.
We conclude with deployment examples, showing how Python scripts can be turned quickly into OR and Prescriptive Analytics applications using amplpy, Pandas, and the Streamlit app framework. Deployments are supported on traditional servers and in a variety of modern virtual environments including containers, clusters, and cloud machines.
Location: CC-North 121 B
Exploring the use of Topological Data Analysis in Image, Time-series, Text, and Social Media
Presented by: Edward Kibardin
Using DataRefiner, we’ll tap into Topological Data Analysis (TDA) and deep learning methods. Together, we’ll navigate data from images, time series, and texts. We’ll spot complex trends in the data using clustering and identify the links between different aspects of the data using cause discovery techniques. Join to see the TDA analysis of an enormous amount of Reddit comments – 2,5 billion to be exact.
Location: CC-North 121 C
Application-building with GAMS: Model Development, Data Transfer, and Deployment
Presented by: Atharv Bhosekar
The General Algebraic Modeling System (GAMS) allows modelers to create optimization-based decision support applications. In this workshop, our first focus will be on model development with GAMS. We will explore what a model entails, how to solve different problem types (linear, mixed-integer, non-linear) using GAMS, as well as how to switch solvers and separate the model code from input data using GDX. Additionally, we will demonstrate how a GAMS model can be integrated and transformed into an effective application. An essential step in this process is ensuring efficient data transfer. To achieve this, we will showcase the use of the embedded code facility, GAMS Transfer API, and tools like GAMS Connect. Lastly, we will introduce the GAMS Engine, a powerful tool for solving GAMS models either on-premises or in the cloud.
Location: CC-North 122 A
Building Sustainable Supply Chain and Logistics: The Case of JD.com
Presented by: Hau Lee, Yongzhi Qi, and Ningxuan Kang
JD.com is China’s largest retailer and a global leader. By operating 1500+ warehouses, and fulfilling 90% of customer orders same or next day, JD.com have achieved a very high automation rate end-to-end. As a company also focused on social responsibility, JD.com has made great effort to establish a supply-chain end-to-end mobile carbon emission footprint management platform, which enables JD.com to work better on carbon emissions for its customers and to identify greenhouse gas decarbonization opportunities for e-commerce scenarios. This speech will help you to uncover the potential supply chain sustainability using machine learning, operation research, AI and IOT techniques with real-world cases.
Location: CC-North 122 B
Performing Basic-to-Advanced Statistical Analyses in an Interactive “No Code/Low Code” Environment
Presented by: Kevin Potcner
A statistical scientist from JMP will demonstrate a wide collection of platforms in JMP software for analyzing data and teaching data science in a ‘no code’ interactive environment.
Topics will include:
- Importing and preparing the data for analysis
- Discovering and describing features in univariate and multidimensional data through interactive visualizations
- Classic Inference and estimation
- Designing and analyzing research experiments
- Conducting comparative statistical analyses
- Building statistical models for describing relationships and predicting outcomes
- Analyzing unstructured text data
- Preparing graphs for publication submission
Location: CC-North 122 C
De-risking Decision Model Rollout with Batch, Acceptance, and Shadow Testing
Presented by: Sebastián Quintero
Isaac Asimov once said, “The most exciting phrase to hear in science, the one that heralds new discoveries, is not ‘Eureka!’ but ‘That’s funny…’” Asmiov’s words speak to the opportunities, curiosities, and showstoppers that surface while building and deploying decision algorithms — but ideally not before an algorithm impacts production systems. That’s where model testing comes in. Testing is a useful tool for algorithm development teams to avoid expensive mistakes, troubleshoot bugs, establish a repeatable path to production, and give stakeholders confidence in results. How does a new model perform compared to an existing one? How do solutions change as constraints are added or removed? Does the model perform as I’d expect before deploying to production? In this workshop, we’ll explore these questions in addition to walking through the different types of model testing, a framework for bringing them all together, and hands-on examples for how to use and incorporate an optimization model testing toolkit into your OR operations.
Location: CC-North 123
Introduction to WordStat Text Analytics
Presented by: Normand Peladeau
Provalis Research CEO Normand Peladeau will identify the main challenges in the analysis of unstructured text data and demonstrates how Provalis Research’s text analytics software WordStat can help you deal with those challenges and analyze large amounts of text data using exploratory text-mining and quantitative content analysis. He will show you how you can quickly and easily explore very large amounts of text data with advanced topic modeling and other techniques as well as how to build and use dictionaries for automated content analysis.
Location: CC-North 121 A
A Real-time Collaborative and Cloud-based Open-source Ecosystem with AI
Presented by: Blaec Bejarano
This workshop serves to highlight a few unique capabilities that CoCalc possesses which enhance the collaborative performance of computational science. This includes a Jupyter integration with OpenAI, computational whiteboard environment (and slideshow mode), version control via TimeTravel, and more.
Location: CC-North 121 B
The New Xpress Global Optimization Solver and Other Advances in Optimization
Presented by: Michael Perregaard, Susanne Heipcke, Jeff Day, and Robert Ashford
This workshop will 1) introduce the new Xpress Global Solver, 2) cover the next generation of features in the Mosel programming language, 3) discuss the ODH MIP accelerator package from Optimization Direct), and 4) describe how Xpress Insight allows you to put the power of your analytical models in the hands of business users in far less time than traditional tools.
- The new global optimization engine in the FICO Xpress Solver can solve mixed-integer, nonlinear optimization problems to proven global optimality. Building upon Xpress Nonlinear and its multi-start extension, Global Optimization offers a new solution methodology for nonlinear problems, supporting all kinds of algebraic nonlinearities (powers, exponentials, trigonometrics), in addition to all of the discrete entities that the Xpress Optimizer supports (binaries, integers, indicators, special-ordered sets, piece-wise linear functions).
- Mosel 6 is a major milestone that reinforces the programming features of the Mosel language. It allows for a more flexible design, implementation, and version management of Mosel packages. Mosel addresses the needs of large-scale application development that we and our clients also demand for non-optimization-based applications. It supports testing systems, debugging functionality, and long-term compatibility management for user and system components. Besides the new programming features, the Mosel 6 series also provides numerous extensions to Mosel’s optimization capabilities. We will provide examples that illustrate how to use the new Global Solver with Mosel, how to formulate and solve optimization problems with multiple objectives, and extensions to the IIS (Irreducible Infeasible Set) functionality. We will also briefly cover new versions of Mosel open-source components, including new features of the online-doc generation, new data structures, and support for testing.
- ODH is a software package from Optimization Direct Inc. that runs concurrently with MIP optimizers like FICO Xpress to accelerate and extend their capabilities for large and difficult optimization models. Using thread synchronization and heuristics like decomposition methods, ODH generates and provides good solutions to the MIP solver, allowing it to converge faster. In tests by Optimization Direct, ODH reduced optimality gaps by 40% on mid-sized models and solved previously unsolved instances from MIPLIB. It is used by customers in scheduling, telecom, retail, logistics, etc. to optimize models that would fail without ODH. By leveraging its synchronization and heuristics, ODH enables MIP solvers to tackle larger and harder real-world optimization problems.
- FICO® Xpress Insight is a rapid application development and deployment framework that integrates with Xpress Solver and your own analytics, enables collaboration across multi-functional teams, and deploys decision support or automated solutions in the cloud or on-premises in far less time than traditional application development tools. We will show how you can rapidly convert Python and Mosel models into complete business applications with Xpress Insight to make your analytical models available to thousands of business users.
Location: CC-North 122 B
Build Analytic Models with Help from AI in RASON and Analytic Solver for Excel
Presented by: Daniel Fylstra
This workshop will highlight what’s new in Analytic Solver® for Excel and RASON®, our cloud platform and high-level modeling language for your full range of analytics – optimization, simulation, machine learning, and business rules. See how you can use our integrated, ChatGPT-powered AI Agent, trained on thousands of pages of our guides, to learn new software features and more easily create models.
We’ll show how RASON helps you connect models to cloud data sources, use different model types together in multi-stage decision flows (without writing code), schedule automated data updates and model solves, run models “on demand”, access results from low-code / no-code tools like Power Apps, Power Automate and Power BI, track model versions, runs, and resource usage, and leverage Azure cyber security features to create a modern “zero trust” solution.
As a bonus, you’ll see our patent-pending, fully automated risk analysis of machine learning models – easy to use in RASON, in Excel with Analytic Solver, and in C++, C#, Java, Python and R with Solver SDK®. (If you miss this, check our tutorial session!) See if you agree that this will become the new best practice for machine learning model development and evaluation.
Location: CC-North 126 B
Gurobi 11.0 – Helping You to Build, Solve, and Deploy Optimization Models
Presented by: Zonghao Gu, Ed Klotz, Rodrigo Fuentes, Cara Touretzky, Dan Steffy, and Lindsay Montanari
Regardless of whether you are an experienced mathematical optimization practitioner, a newcomer interested in better understanding the potential of mathematical optimization in your organization, or somewhere in between, this workshop has something for you.
Attendees will get a first look at our latest product release, Gurobi 11.0—including exciting performance improvements to Gurobi’s core algorithms and additional support for nonlinear functions. The latter includes direct support of general nonlinear functions beyond the piecewise linear approximations provided in previous versions. The workshop will also include an interactive panel discussion “From skepticism to success: getting buy-in for optimization projects” with optimization experts across academia, industries, and Gurobi.
The Gurobi team will also be presenting a live demo of OptiMods, an open-source project that provides Python users with high-level access to optimization capabilities. Developed by the Gurobi Tech teams, the project aims to put optimization in the hands of more people—especially those without prior knowledge of optimization and mathematical modeling. While primarily intended for newcomers, experienced optimization practitioners may gain insights on obtaining acceptance of optimization-based solutions.
All Workshop attendees will receive a free Gurobi t-shirt.
Location: CC-North 121 C
Littlefield 2.0 — A New Version of the Online Game for Operations Management Courses
Presented by: Samuel Wood
After 24 years there is a new version of Littlefield! Littlefield is a competitive online simulation of either a factory or a medical laboratory that has been by more than half a million students in 500+ universities in 60+ countries to excite and engage students in operations management topics like process analysis and inventory control. This presentation will introduce a newly updated version 2 of the game that will go into production in 2024. Participants are encouraged but not required to bring a laptop.
Location: CC-North 122 A
SAS Analytics and Digital Twins
Presented by: Rob Pratt, Hossein Tohidi, and Bahar Biller
SAS offers extensive analytic capabilities, including machine learning, deep learning, natural language processing, statistical analysis, optimization, and simulation. SAS analytic functionality is also available through the open, cloud-enabled design of SAS® Viya®. You can program in SAS or in other languages – Python, Lua, Java, and R. SAS Analytics is equipped with AI-enabled automations and modern low-code or no-code user interfaces that democratize data science usage in your organization and offer unparalleled speed to value.
We’ll first review the SAS analytics portfolio, then highlight recently added optimization features, and finally explore case studies in optimization, network analytics, and simulation, with an emphasis on projects involving digital twins. The concept of digital twin plays a paramount role in the digital transformation efforts of our customers. We will describe how we envision digital twins, share our development framework, and highlight SAS enabling technologies and differentiators. We will further share our experience of building digital twins and conclude with an overview of SAS customer outcomes.