Just as the emergence of steam engines and the telegraph changed the world during the industrial revolution, artificial intelligence technologies such as machine learning are reinventing the way we live and work. Engineers and organizations alike must adapt or risk becoming obsolete.
Machinery.ai is a community and event series where developers come together to explore the digital machinery of the new era.
At our upcoming conference, experience a day of learning from industry experts to look “under the hood” at the building blocks of machine learning.
Modev is committed to creating diverse and inclusive events and encourages people to participate regardless of gender identity or expression, age, family or marital status, national origin, physical and mental ability, race, ethnicity, religion, socio-economic status, veteran status, or sexual orientation. We believe the best events are ones where all types of people feel welcome and included, and are represented in both the audience and the speakers. Please join us!
Ethan Chumley is a Campaign Technology Advisor and Tech Evangelist for Microsoft’s Tech & Corporate Responsibility team. A hands-on technologist, he has worked on tech, data, and security strategy in the political and civic spaces, most recently focusing heavily on the 2016 Presidential cycle. Ethan works across a broad suite of Microsoft technology, including “big data” analytics, visualizations, Azure, and Office 365. Formerly, Ethan was a field consultant in Microsoft Consulting Services (MCS), specializing in the Internet of Things and IoT analytics. Prior to Microsoft, he worked in energy modeling for an Engineering firm in New York City. Ethan received a B.S. in Engineering from Cornell University, grew up near Denver, CO, and is an avid skier.
With a BS in Applied Mathematics, and a MS in computer science, Anna launched her career in engineering in 2002 working on the Thirty Meter Telescope project. Over the next 12 years, she specialized in remote sensing algorithms, culminating as the principle investigator in an Office of Naval Research contract on the classification of signals. In 2014 she took her breadth of knowledge in applied research to the IBM Watson group. Within IBM Watson she has defined and measured key cognitive metrics necessary to track the improvement and value of the machine learning training cycle, and continues to be passionate about quantifying and improving systems that can improve people’s lives.
Ensemble Models in Machine Learning are a bit of a polarizing topic. People either love them, or hate them. A common criticism is that they are not mathematically rigorous. However, nothing could be further from the truth, as one of the earliest foundations of Machine Learning was in form of theoretical validity of ensemble models. Nevertheless, there are noticeable differences between theory and practice when it comes to ensemble models. This talk will attempt to compare the theory and practice and present suggestions to narrow the gap between the both.
Overview, prerequisites, and implementation details of getting started with Learning To Rank. Why relevancy tuning will be taken over by machine learning, but not yet.
With so many options to choose from how do you select the right technologies to use for your machine learning pipeline? Do you purchase bare metal and hire a devops team, install Spark on EC2 instances, use EMR and other AWS services, combine Spark and Elasticsearch?! Attend this talk to hear first-hand experience of building ML pipelines: what options were looked at, how the final solution was selected, the tradeoffs made and the final results.
Machine learning has delivered great results in domains where data is easily available. But what about real-world domains where large amounts of data are hard to find? Often, the industries that could benefit from AI and ML most - energy, finance, medicine - are also most constrained in their data availability and access. What can we do when finding large amounts of labeled data to train with is near impossible? What if all we have is small data? How can we still apply ML techniques to these domains? In this talk, I will present an overview of approaches like transfer learning, adaptive sampling and semi-supervised learning and show how these techniques can shine a ray of hope in a barren data landscape. :)
Using CNTK Train Reinforcement Learning Model play game with GPUs in Azure
The output of machine learning system can alwaysbe improved. Better training data, algorithms more suited to your use case, and system improvements based on threshold setting can all be employed.However, you will find that after each iteration, the system will improve less and less...much like the radius of a spiral as it makes rotations around the origin. In this talk, I will describe the process to improve two different types of problems: sentiment detection and question answer.
Analytics, machine learning, data models and neural networks are valuable tools, but customers buy functionality and usability. Deriving and delivering value from data requires dev teams to consider far more than the efficiency of their code and efficacy of their algorithm. This session will highlight key capabilities that can make the difference between a successful machine learning product and interesting "science project".
This talk ultimately focuses on the ways in which Semantics and Linked Data are an essential part of Industry AI solutions for Big Data in the future, illustrated with examples and case studies (language interpretations, lifting Relational DB schema, sentiment analysis, automation, IoT, and other knowledge engineering areas). This discussion includes a brief introduction into the field, how data can be effectively used in a semantic framework, how building context awareness from a data perspective can be useful and exciting, and most importantly how Semantics/Linked Data are not solutions in and of themselves to reach true AI, but rarely are analytics truly intelligent without them.