Tech Talk

Events Search and Views Navigation

Event Views Navigation

Today

Latest Past Events

Tech Talk: UX Concepts for Generative AI Experiences

Online

What should be considered when AI provides or mediates user experiences? We'll discuss the key principles and heuristics that can guide our approach and explore the mindset necessary for AI-driven user experiences (UX). About Karl Mochel Karl Mochel is a user experience architect with 20+ years of experience in his field. Like an architect designs physical space, a user experience architect designs devices and software to create experiences that are useful and desirable or efficient. He has done this for Oracle, Google, Autodesk, VMware, and GE, as well as other notable companies. Karl's work with Artificial Intelligence (AI) began at GE Global Research, where he worked on IoT analytics for fleets of industrial devices such as jet engines and wind turbines. As generative AI has grown in relevance, he looks at how it will impact businesses and what activities need to be considered to rise with Gen AI, not be overwhelmed by it. Register Online: https://forms.gle/j2EtKLzA9896inLZ7

Free

Tech Talk:Deep Symbolic Optimization: A framework for symbolic optimization

Online

Today's Talk We propose a framework to optimize hierarchical, variable-length objects under a black-box performance metric, with the ability to incorporate in situ constraints. The framework uses deep learning via a simple idea: use a large model to search the space of small models. Specifically, we use a neural network (NN) to emit a distribution over tractable discrete objects and employ a novel risk-seeking policy gradient to train the NN to generate higher scoring objects. As an application, we discuss the problem of symbolic regression, i.e., discovering the underlying mathematical expression that fits a dataset. The resulting algorithm outperforms several baseline methods in its ability to exactly recover symbolic expressions on a series of benchmark problems, both with and without added noise. We also discuss the hybridization of our framework with other modern and classical approaches in symbolic regression, including evolutionary algorithms and linear methods. Finally, we present applications in the reinforcement learning setting, where we use our framework to discover simple and interpretable control policies. The discovered symbolic policies outperform highly tuned NN-based policies, with the additional benefit of being interpretable and verifiable.   Mikel Lanajuela, PhD Mikel Landajuela is a machine learning researcher at Lawrence Livermore National Laboratory (LLNL) in Computational Engineering Directorate. He completed his Ph.D. at Université Pierre et Marie Curie and Inria, for which he was awarded the SMAI-GAMNI award 2017 (French Society of Industrial and Applied Mathematics) for the best thesis in numerical methods for the mechanical and engineering sciences. He has hold postdoctoral appointments at Politecnico di Milano in Italy and LLNL. Recently, his work in symbolic optimization using deep learning has led to several publications in top-tier machine learning conferences and several recognitions, including the “Interpretable Symbolic Regression for Data Science” competition-award at GECCO 2022.

Free

Tech Talk: ATOM Modeling PipeLine (AMPL) for Drug Discovery

Online

Register Online - FREE Tonight's Talk The ATOM Modeling PipeLine, or AMPL is an open-source, modular, extensible software pipeline for building and sharing models to advance in silico drug discovery. One of the key requirements for incorporating machine learning (ML) into the drug discovery process is complete traceability and reproducibility of the model building and evaluation process. AMPL was developed with this in mind as an end-to-end modular and extensible software pipeline for building and sharing ML models that predict key pharma-relevant parameters. AMPL extends the functionality of the open source library DeepChem and supports an array of ML and molecular featurization tools. In this talk, results of extensive benchmarking on a wide variety of pharmacokinetic and safety data sets will be presented, with an exploration of the effects of different featurization and model types on model accuracy. Hiranmayi Ranganathan, PhD Hiranmayi is a machine learning specialist at Accelerating Therapeutics for Opportunities in Medicine (ATOM). As part of the data modeling team, she works on building deep learning models of secondary pharmacology, with the goal of predicting adverse effects of drug candidates before they advance to animal and human trials. She joined Lawrence Livermore National Laboratory (LLNL) in July 2019 and has been part of the machine learning group since then. Before that, she did her Ph.D. in Electrical Engineering from Arizona State University. Her research interests are in Deep Learning, Active Learning, Emotion Recognition, and Deep Learning for drug discovery. Community Partner Event This event is a cross-over event to make the attendee's aware of the REWORK MLOps Trusted AI Summit 2022 on June 15-16 and provide an exclusive preview of the talks that will be featured at the summit. The MLOps Trusted AI Summit is a collaborative event for Artificial Intelligence and Machine Learning. The summit already has confirmed speakers from companies such as Chick-fil-A, Uber, Capital One, Meta, Twitter, and many more! ALL attendees of this Tech Talk will be given a code to sign up for the summit with a 25% discount code and we will raffle out 5 free tickets (a $2,295 value) for all present attendees that are also IEEE members!  

Free