AI for Good Summit schedule

Towards a Sustainable Food Supply Chain Powered by AI

About 30-40% of food produced worldwide is wasted. This puts a severe strain on the environment and represents a $165B loss to the US economy alone. This talk explores how AI can automate decisions across the food supply chain in order to reduce waste and increase the quality and affordability of food. We focus our attention on supermarkets, and describe an intelligent decision support system that uses model-based reinforcement learning to help store managers make better operating decisions. In a real-world pilot with a US supermarket chain, our system yields significant waste reductions, while increasing store profits.

Volodymyr Kuleshov is the co-founder and CTO of Afresh, a startup that uses AI to fight food waste. He obtained his Ph.D. in Computer Science from Stanford University, where he was the recipient of the Arthur Samuel Best Thesis Award. Volodymyr’s work has been featured in Nature Biotechnology, Nature Medicine, Nature Communications and Scientific American and was awarded an NSERC Post-Graduate Fellowship and a Stanford Graduate Fellowship. He is also a co-inventor of the machine learning technology that powers the phased genome sequencing product of Illumina Inc, following its acquisition of the Stanford spin-off Moleculo.

Climate Zones Classification for Agriculture Applications using Machine Learning

Modern climate classification techniques typically employ intra-annual variability of climate data--e.g. temperature and precipitation--in order to segment geographical regions into meaningful zones with similar physical characteristics. For instance, common classification schemes like the Koppen-Geiger rely on human-defined clusters. Here we show a generalizable method for creating meaningful climatically-relevant zones tailored to each user’s needs. To accomplish this, we first explored several clustering methods by comparing performances and weighing potential advantages and/or disadvantages to each technique. We found out that our streamlined clustering procedure can benefit the real state and agricultural sectors especially, by taking into consideration climate change conditions.

Enthusiastic, curious, and with multidisciplinary interests. I want to help bridging the gap between science, engineering and the society's interests. I particularly enjoy using novel data science techniques in conjunction with machine learning algorithms to extract knowledge from environmental Big-Datasets, like global climate models, weather station networks and gridded observation-based products. Overall, I have experience with predictive analytics, forecasting, time-series analysis, nonlinear methods and optimization problems.

AI Applications for Earthquake Monitoring

Diverse algorithms have been developed for efficient earthquake signal detection and processing. These algorithms are becoming increasingly important as seismologists strive to extract as much insight as possible from exponentially increasing volumes of continuous seismic data. Convolutional and recurrent networks have each been shown to be promising tools for this. We have developed a number of deep learning tools for more efficient processing and characterizing of earthquake signals. In our presentation, we demonstrate the performance of some of these tools applied to seismic signals. AI-based techniques have the potential to improve our monitoring ability and as a result understanding of earthquake processes.

Mostafa Mousavi received the M.S. degree in Risk Engineering from the University of Tehran, Tehran, Iran in 2010 and the M.S. and Ph.D. degrees in geophysics from University of Memphis, TN, USA in 2017. He is currently a Postdoctoral Research Fellow at Stanford University, CA, USA. He is the author of one book, 20 journals, and 3 conference papers. His research interests include machine learning/deep learning, signal processing, statistical seismology, and observational earthquake seismology. Dr. Mousavi is a fellow of the National Elite Foundation of Iran and a recipient of SEP/SEG award by ExxonMobil in 2014.

Machine Learning for Environmental Sustainability

The advances in machine learning are revolutionizing how we understand and design effective strategies to deal with forest and other natural resource sustainability challenges. In this talk, I will show how state-of-the art machine learning algorithms advance analysis of the social and ecological impacts of natural resource policies and generate evidence to prioritize governance actions to further environmental sustainability goals. I will also explore reasons for slow adoption of automated decisionmaking in natural resource management and conclude with possible directions for mainstreaming artificial intelligence and machine learning in forest conservation and management.

I am an interdisciplinary environmental social scientist interested in evaluating and predicting the impacts of forest conservation policies at local and global scales to build knowledge on sustainable Social-Ecological Systems (SESs). I am particularly interested in exploring use of artificial intelligence and machine learning approaches in environmental sciences especially in the fields of policy impact evaluation, forest conservation and management. I have also experience in using advanced spatial analytical techniques, machine learning and causal inference based statistical approaches to explore research objectives in the field of environmental and forest governance.

Deep Learning for Climate Science

Climate change is one of the most important problems facing humanity in the 21st century. Climate simulations provide us with a unique opportunity to understand the evolution of the climate system subject to various CO2 emission scenarios. Unfortunately, large scale climate simulations produce 100TB-sized spatio-temporal, multi-variate datasets, making it difficult to conduct sophisticated analytics. In this talk, I will present our results in applying Deep Learning for semi-supervised learning of extreme weather patterns. I will conclude with a summary of current applications of Deep Learning for a broad set of scientific use cases, and open research challenges for the future.

Prabhat leads the Data and Analytics Services team at NERSC. His current research interests applied statistics, machine learning, and high performance computing. He has worked on topics in scientific data management, parallel I/O, scientific visualization, computer graphics and computer vision in the past. Prabhat received an ScM in Computer Science from Brown University (2001) and a B.Tech in Computer Science and Engineering from IIT-Delhi (1999). He is currently pursuing a PhD in the Earth and Planetary Sciences Department at U.C. Berkeley.

Session takeaways: 1) An introduction to DRL models, algorithms and techniques; 2) Examples of DRL systems; 3) How DRL can be used for practical applications.

This session is for C-level and Senior Level Executives.

Moogsoft develops AIOps technology that helps enterprise ITOps and DevOps teams become faster, smarter and more effective. Moogsoft AIOps’ real-time machine learning algorithms help teams remediate issues that impact their customers’ experience by: 1) Reducing operational noise (alert fatigue) across your production stack. 2) Proactively detecting incidents and correlating events across your monitoring ecosystem. 3) Streamlining collaboration and workflow across teams and toolsets. 4) Codifying knowledge to make operators smarter when encountering future Incidents.

Session takeaways: 1) Evaluate what is the purpose of the framework - what do you want to accomplish? 2) The strenghts and weaknesses of common frameworks; 3) What programing language will be used to develop models?

This session is for data scientists and machine learning engineers.

Session takeaways: 1) Why is data preparation important? 2) Where should you gather the data; 3) How do you handle missing data

This session is for everyone.

Session takeaways: 1) What is emerging in the field? 2) What industry do the next generation believe will be most impacted by AI over the coming years? 3) What inspired these rising stars to explore STEM and what is the role of all stakeholders in encouraging young people to do the same?

IBM Services partners with the world's leading companies to reimagine and reinvent to build smart business -- from end to end and from both the outside in and the inside out. Our clients are optimizing processes with emerging technologies to deliver more intelligent workflows as an integral part of transforming into a cognitive enterprise. We partner with clients to design solutions, modernize and optimize their business, deliver sustained value and empower their people. Our end-to-end capabilities take clients from inception to support and are backed by the power of the full IBM portfolio. A new era for business has arrived. ibm.com/services/process

The session is part of the Decision Maker Access pass. View further information on this here.

Session Takeaways: 1) Discover how to accelerate ML initiatives in your business; 2) What does Artificial Intelligence and Machine Learning assisted digital transformation look like in practice, and what are the benefits of adoption?; 3) Design a framework on where to start.

The session is for C-level and senior executives - responsible for driving digital transformation, innovation and introducing new tech methods and tools.

The session is part of the Decision Maker Access pass. View further information on this here.

Session takeaways: 1) What are some of the main short, medium & long-term issues in integrating AI into healthcare & financial sectors? 2) What learnings can be shared cross-industry? 3) What are the top benefits for integrating AI into each sector?

Session takeaways: 1) What are the short, medium and long-term challenges in investing in AI to solve important problems in society? What are the main success factors for AI startups? 3) What are the challenges from a VC perspective?

Session takeaways: 1) What are the key ethical questions that should be considered for AI implementation/projects? 2) How should we determine the success and fairness of AI? 3) How can we work through challenges & issues in a collective and inclusive manner?

Session takeaways: 1) What do we mean by accessible AI and why is it important? 2) How can we as individuals and companies play our part to increase the accessibility of AI; 3) How can an increased AI reach impact the application of AI?

Session takeaways: 1) Is AI the answer to solving this challenge, and if so, why? 2) What partnerships need to be formed to move this problem forwards? 3) What is the role of industry and academia to solve this challenge in question?

Session takeaways: 1) An overview of promising open source tools 2) What are the challenges associated with training RL agents? 3) Building scalable frameworks

This session is for: Data scientists, ML Engineers, RL Engineers