FREE | ONLINE | CONFERENCE
an IBM Community Virtual Event
October 11th, 2018
Join developers and their advocates as they talk about projects and technologies they contribute to and depend upon
AI's Elephant in the Room
AI is driven by data. Where is this data? Is it the right data for the situation and how do we ensure the AI remains true to its purpose as it learns from new incoming data? In this talk, we will look at the structure of the data ecosystem needed to feed AI and review how the industry efforts around the Egeria Open Source project are aiming to support it.
Master Inventor, Fellow of the Royal Academy of Engineering
Building a Secure and Transparent ML Pipeline Using Open Source Technologies
The application of AI algorithms in domains such as criminal justice, credit scoring, and hiring holds unlimited promise. At the same time, it raises legitimate concerns about algorithmic fairness there’s now a growing demand for fairness, accountability, and transparency from machine learning (ML) systems. And we need to remember that training data isn’t the only source of possible bias and adversarial contamination. It can also be introduced through inappropriate data handling, inappropriate model selection, or incorrect algorithm design.
What we need is a pipeline that is open, transparent, secure and fair, and that fully integrates into the AI lifecycle. Such a pipeline requires a robust set of bias and adversarial checkers, “de-biasing” and ""defense"" algorithms, and explanations. In this talk we are going to discuss how to build such a pipeline leveraging open source projects such as AI Fairness 360 (AIF360), Adversarial Robustness Toolbox (ART), and Fabric for Deep Learning (FfDL), and Seldon
Sr. Technical Staff
AI and Deep Learning Platform
No Data, No AI
AI is rapidly growing, being used to increase sales, make better decisions faster and even potentially safe lives through combining data from a lot of different sources. AI is build with models, but these machine learning or deep learning models are only as good as the data they are trained on. What is good data, and is more data better data? In this talk we will explore what good data is, how to turn bad data into useful data and how to store and access the data when it grows into big data with practical examples.
Model Asset Exchange
We've all heard that AI is going to become as ubiquitous in the enterprise as the telephone, but what does that mean exactly? In this talk, we'll break down the challenges a domain expert faces today in applying AI to real-world problems. We will talk about the challenges that a domain expert needs to overcome in order to go from "I know a model of this type exists" to "I can tell an application developer how to apply this model to my domain." We'll conclude the talk with a live demo that showcases how a domain expert can cut through the stages of model deployment in minutes instead of days using the IBM Developer Model Asset Exchange.
Chief Architect, Center for Open-Source Data and AI Technologies
IBM Fellow, AI Science
Ethics in AI
While the landscape of artificial intelligence technology is making groundbreaking advances, we must find ways to ensure that scientists and developers alike are using this technology to make ethical decisions. Aleksandra (Saška) Mojsilović is a scientist, Head of AI Foundations at IBM Research, Co-Director of IBM Science for Social Good, and IBM Fellow. She is a Fellow of the IEEE and a member of the IBM Academy of Technology. We’ll talk with her take on the current landscape of ethics in AI, and where she sees AI going in the future.
Hear from IBM product experts who will discuss the latest AI solutions and strategies that you can instantly translate to your business
Architect, Senior Technical
Staff Member for
Data Science Experience
Develop and deploy a real-time ML application on IBM Cloud Private for Data
Machine Learning has changed the way businesses operate, and to truly get the benefit of Machine Learning, especially for real-time situations, enterprises need to ensure high throughput and reliability from all components in their solution. Such applications need to plan for extreme compute elasticity & high availability and strategies for dealing with outage free rolling upgrades to be truly successful, i.e. - enterprises need applications that exhibit cloud like characteristics in their own data center private clouds, behind their firewalls. So, how does one go about doing this without the full infrastructure of a Cloud IaaS & Paas at their disposal?
In this session for developers and data scientists, we will walk through a simple, but non-trivial, example of building a cloud-native real-time application that invokes a machine learning model for predictions. We will discuss how ICP for Data can help you train models & deploy scalable scoring services as well as how the Kubernetes based platform can also host your cloud native application & ensure reliability and scale.
Advisory Software Engineer
Building Custom AI Web Services
Artificial Intelligence turns insights from your data into recommended actions. The AI process consists of a few key steps: preparing data, building and training a model, deploying the model as web services and retraining the model with new data.
Applications are relying on model predictions to forecast a future behavior. As the incoming data changes its distribution over time, predictions accuracy can suffer. To tackle this challenge, the model needs to be retrained and redeployed. Models can be retrained manually or automatically. Retraining and redeploying the model unlocks continuous learning which is key to AI and AI-based web services.
This session will explore how to build ready to consume, custom AI based web services.
Executive IT Specialist
IBM Hybrid Cloud
Governed Data Science
Move to a digital enterprise requires that we collect more data and use it to help serve our customers better. Whilst this is a great opportunity, it also comes with increased responsibility on how we handle our data/analytics and remain compliant with regulatory requirements. Data science has been rapidly growing over the past few years with organizations using new tools, technologies, algorithms to leverage historical data and make smarter decisions. In this talk, we will walk through the key aspects of establishing a governance framework, the need for governance and how we can apply it in the context of data science.
Db2 Event Store
Fast Data Ingest and Analytics
In this session, we will review an end to end example running on the new IBM Fast Data Platform. Combining multiple open source services: Kakfa, Spark or Grafana, we will expose how IBM and Lightbend joint solutions provide a complete toolchain for Java and Scala developers. You will learn how to easily build and deploy AI and cognitive applications in both on-premises or cloud-based environments. The IBM Fast Data Platform brings together Lightbend's Fast Data Platform, Db2 Event Store & IBM Data Science Experience Local to enable a new era of event-driven business insight and opportunity. This session is the perfect place to learn how to manage data in motion for your streaming applications.
IBM PowerAI Team
Infrastructure to AI
This is an overview of IBM’s PowerAI ecosystem. PowerAI makes deep learning and machine learning more accessible to your staff, and the benefits of AI more obtainable for your business. It combines popular open source deep learning frameworks, efficient AI development tools, and accelerated IBM® Power Systems™ servers. Now your organization can deploy a fully optimized and supported AI platform that delivers blazing performance, proven dependability and resilience. IBM PowerAI Enterprise is a complete environment for data science as a service, enabling your organization to bring new applied AI applications into production. We will also spend time discussing client use cases and best practices.
Putting AI to Work: Extending Beyond the Lab
Businesses everywhere are exploring AI’s potential. But it’s not as simple as deploying traditional software. Businesses need to overcome a variety of hurdles to scaling their AI, from incorporating new skills and tools, to establishing new methodologies and overcoming a lack of visibility into the decisions AI systems produce. As AI deployment accelerates, compliance and trust are critical. We’ll discuss lessons learned from thousands of AI engagements across industries and how to foster collaboration between the teams that operate AI and the users of these applications.
AI & ML
Automatic ML infused classification for Governance and GDPR compliance
Most organizations spend a great deal of time and energy wrestling dirty, poorly-integrated data. They either cannot find the right data or cannot trust the data they find. On top of that, they must deal with multiple regulations in their industry that are barriers to self-service and data democratization. As a result, they try to fix their data through a variety of labor-intensive tasks, from writing custom programs to global replace functions—overall diminishing their productivity as data analysts and data scientists. This talk will take you through the journey of data curation and governance using AI and machine learning at enterprise scale.
AI & ML
Watson Studio: An Extensible Environment for Self-Service AI
Watson Studio is a rich platform for wide range of analytics, catering to many kinds of users, including programmers, statisticians, data engineers, and business analysts. While it has many embedded tools, Watson Studio was designed to be extensible. This talk will cover many of these extensions, including supported data stores, data modeling tools, data governance tools, and visualization engines. While there is a concern in many organizations that self-service analytics and data science efforts require new tools that are disconnected with existing data infrastructure, Watson Studio is designed to complement it.
Learn how IBM Community Partners can help you tackle your AI & Data Science challenges
Alyssa Simpson Rochwerger
Empowering Data Scientists to Train, Test and Tune AI for the Real World
In this insightful session, see how IBM Watson Studio users will be able to label their data using Figure Eight's Human-in-the-Loop Machine Learning platform directly from their Watson Studio project. Figure Eight's enterprise-grade platform reduces the time you spend wrangling data and produces high-quality training data for your machine learning models.
An Agile Data Preparation and Exploration Process for AI and Machine Learning
While AI and machine learning offers even deeper insights than ever, data science processes are hindered by data science teams spending 80% of their time preparing, exploring and managing data. The most time-consuming aspect is not cleansing but rather getting access to the right datasets, determining if a data can contribute to the model, feature engineering for the model, and validating model results in detail.
Join us for this session where we will examine how to create an agile, iterative data preparation and exploration process using the right platforms and tools. You’ll learn how to use advanced data prep and exploration features across the entire data lifecycle to speed up AI and machine learning analytic cycles.
Teaching AI Skills at Udacity
Artificial intelligence is a dynamic field, and demand for people with AI skills is growing in industry and academia!
In this talk, you'll learn about how Udacity aims to teach in-demand AI skills. Cezanne Camacho, curriculum lead for the Computer Vision and Deep Learning Nanodegree programs at Udacity, will discuss approaches to learning AI skills, project-based learning, and the programs and resources that Udacity offers. At the end, you'll learn the basics of neural networks, and we encourage you to share this knowledge with others!
Co-Founder & CEO
How to Select Your Trusted Partner in Data for AI
High-performing machine learning models require large quantities of quality training data to execute properly. Getting that data is a complex, multidisciplinary process that’s often as time-consuming as it is tedious. Currently, data scientists spend 80% of their time cleaning and structuring data. In order to maximize time spent focused on model building those researchers turn to partners to help fulfil their data needs. However, this process can be risky. Most of the time, the source of the data is unknown, the process by which it is sourced is non-transparent, and the data itself still needs to be redone or fixed. That means lower-quality data is often fed into models, lessening ROI and potentially damaging brands as a result. In this talk, we will identify what data scientists should look for when considering a data partner, namely guaranteed exit criteria in the SLAs, the level of domain specialization of the partner, and international multilingual support.
Sr. Data Scientist
Jump Start AI at your Organization
The most enticing opportunity for the BI function is to pivot into data science and Artificial Intelligence. With the support of an experienced data scientist, it is possible to capitalize on your strong organizational networks and deep cross-functional, institutional, and business process knowledge to enable and empower AI in your organization. Attend Ironside’s session to:
• Understand your AI opportunity
• Learn tips on developing and executing a plan
• Learn best practices to operationalize and optimize AI
Develop an AI roadmap today and start your journey to grow revenues, maximize efficiency, and understand and minimize risk in your organization.
Text Modeling Using Tidy Data Principles
Text data is increasingly important in many domains, and tidy data principles and tidy tools can make text mining easier and more effective. In this talk, learn how to manipulate, summarize, and visualize the characteristics of text using these methods and R packages from the tidy tool ecosystem. These tools are highly effective for many analytical questions and allow analysts to integrate natural language processing into effective workflows already in wide use. Explore how to implement approaches such as topic modeling and building text classification models.