Innovative applications using Deep Learning and Blockchain w/ Insight AI Fellows

A panel of Insights Fellows, sharing their projects in Deep Learning / AI.

Speaker 1 – Michelle Bonat, CEO & Co-Founder, Data Simply, Inc –
“What I Learned Predicting Crypto Flows with Deep Learning and Blockchain”

Speaker 2 – Josh Deetz, Physical Data Scientist at Carbon
“Identifying At-Risk Patients using Intensive Care Unit Data”

Speaker 3 – Khyati Ganatra, Data Sceintist at Cequence Security
“Deep Photoshop: Smart logo replacement using image infilling”

6:30 arrive, register, pizza, network
7:00 ACM intro
7:10 to 8:30PM – Talk by 3 speakers followed by Q&A.
7:10 Live stream Click the link to set your PC up and get ready
https://www.youtube.com/watch?v=

Abstracts:
Deep learning can seem like a dark art. The reality is that it is very achievable to get a model working and predicting well. But deep learning also has some myths and pitfalls of which you should be aware. Michelle Bonat will walk through a project she did to predict cryptocurrency flows using deep learning and blockchain data. This includes code snippets and real world results with various models side by side. I’ll also address myths and preconceptions about deep learning to help you dig in for your own projects.

Josh Deetz will point to today’s hospitals are crowded with patients. Moreover, intensive care units collect lots of data on their patients, from sensors to laboratory tests. However, they have not leveraged all of this information to improve the quality of care. Josh will speak about how machine learning models can use this information to predict patient mortality in intensive care units, and opportunities for the future of healthcare.

Khyati Ganatra will talk about Deep Photoshop, one such use case of using Deep Learning for changing content in an image, like photoshop. Machine Learning and Deep Learning is performing as good as humans in many tasks now. Deep learning, for example, is able to recognize objects in images with better accuracy than humans in real time. Styling the pictures, generating new images, etc. is now possible.

Bios:
Michelle is the CEO / Co-Founder / Data Scientist of Data Simply, Inc.
She runs the business; raised seed funding; automated 99% of ops and pipelines. She is passionate about solving big impactful problems with technology, powered by big data and artificial intelligence. A dynamic hands-on leader, her experience spans data science, engineering, and product management with a track record of success in delivering commercial software products globally at scale. She has successfully run multinational technical teams of 25+ for one of largest tech cos in the world (Oracle), as well as 3 companies which were acquired. Brought products to global market leadership. Shipped technology in more than 25 countries and 30 languages. She is currently CEO/Co-Founder at Data Simply (DataSimply.com) which uses data science to help people invest inline with their beliefs and values (environmental-social-governance), by mining text in company disclosures. There she developed complex real-time data pipelines that have powered more than 100 million automated investment insights. Quantitative Masters degree (MBA, Finance, Kellogg) and hands on AI/machine learning engineer.
LinkedIn: https://www.linkedin.com/in/mbonat/

Josh Deetz, Ph.D., Physical Data Scientist at Carbon currently works as a Physical Data Scientist at Carbon. He has a Ph.D. in Chemical Engineering from UC Davis, where he studied Computational Chemistry. https://www.linkedin.com/in/josh-deetz/

Khyati Ganatra, Khyati is currently working as a Data Sceintist at Cequence Security. She previously worked at IBM in Watson & Cloud Platform. She recently completed her Masters in Computer Science with a focus in Data Science from University of Southern California.
https://www.linkedin.com/in/khyatiganatra/