Cook County

SC19: AI and Machine Learning Sessions Pepper Conference Agenda


AI and HPC are increasingly intertwined – machine learning workloads demand ever increasing compute power – so it's no surprise the annual supercomputing industry shindig, SC19 at the Colorado Convention Center in Denver next week, has taken on a strong AI cast. As we noted recently ("Machine Learning Fuels a Booming HPC Market") based on findings by industry watcher Intersect360 Research, "enterprise infrastructure investments for training machine learning models have grown more than 50 percent annually over the past two years, and are expected to shortly surpass $10 billion, according to a new market forecast," and much of that training calls for HPC-class systems. With that in mind, here's a rundown of AI-related sessions and activities coming up at SC19 (all event locations are in the Convention Center unless otherwise specified): Deep Learning on Supercomputers, 9am-5:30pm, room 502-503-504: This workshop will be led by Zhao Zhang of the University of Texas, Valeriu Codreanu of SURFsara and Ian Foster of Argonne National Laboratory and the University of Chicago and is designed to be a forum for practitioners working on all aspects of DL for science and engineering in HPC and to present their latest research results and development, deployment, and application experiences. Tools and Best Practices for Distributed Deep Learning on Supercomputers, 1:30-5pm, room 201: This tutorial will be led by Xu Weijia and Zhao Zhang of the Texas Advanced Computing Center and David Walling of the University of Texas and is intended to be a practical guide on how to run distributed deep learning over multiple compute nodes. Deep Learning at Scale, 8:30am-5pm, room 207: Led by seven experts from Lawrence Berkeley National Lab, Intel and Cray, this tutorial will focus on the impact of deep learning is having on the way science and industry use data to solve problems and the need for scalable methods and software to train DL models.

Kraft Heinz Appoints New CIO To Deliver An AI Growth Recipe


Kraft and Heinz products are shown on March 25, 2015 in Chicago, Illinois. Kraft Heinz has named a new global CIO as the packaged foods giant seeks to turn around its performance following a sales slide and significant writedowns in the value of some of its most prominent brands. Corrado Azzarita, 49, was previously in charge of IT projects at the company in areas including supply chain, finance, and legal and corporate affairs. He replaces Francesco Tinto, who was appointed global CIO at Walgreens Boots Alliance in September. Kraft Heinz has had a rough year.

PaxeraHealth to show AI-based enterprise imaging at RSNA 2019


PACS/RIS developer PaxeraHealth will highlight its updated PaxeraUltima360 artificial intelligence (AI)-based enterprise imaging software at RSNA 2019 in Chicago. PaxeraUltima360 is designed to use machine-learning technologies to decrease clinician workload by performing basic tasks and improving access to and coordination of patient data and care, the company said. The platform features an AI chatbot called Erabot that facilitates user interaction with the platform, thus speeding up access to relevant patient information. In addition, PaxeraUltima360 can adjust to the preferences of users by monitoring their behavior patterns for certain tasks and providing clinical decision support backed by augmented reading aids, PaxeraHealth said.

Chicago funding news: Artificial intelligence and real estate top recent local investments


Chicago-based legal and compliance company Ascent has secured $19 million in Series B funding, according to company database Crunchbase, topping the city's recent funding headlines. The cash infusion was announced Nov. 5 and led by Drive Capital. According to its Crunchbase profile, "Ascent provides a cloud-based platform that helps financial services firms to keep their businesses compliant. Its platform analyzes business activities, informs about potential compliance obligations and assists in tracking and complying with relevant requirements. Ascent's cloud-based solutions include issue tracking and management, industry reference and research materials and compliance manual documentation."

SIU College of Business celebrates business analytics and artificial intelligence programming launch


In celebration of its newly-launched and highly-anticipated programs in business analytics and artificial intelligence, Southern Illinois University Carbondale's College of Business will host an evening reception for prospective students, business leaders and alumni on Wednesday, Nov. 13, in downtown Chicago. The event will take place from 6 to 8 p.m. on the 27th floor of the Deloitte building (Room 27E047), located at 111 S. Wacker Drive. Attendees will have the opportunity to meet faculty teaching these innovative courses, as well as analytics industry executives serving on the board of the university's one-of-a-kind Pontikes Center for Advanced Analytics and Artificial Intelligence. SIU recently launched an Analytics Concentration for its nationally-ranked online MBA program, a Bachelor of Science in Business Analytics, and will soon introduce a full graduate analytics program. All of these programs are uniquely designed to bridge the gap between data science and business by arming the managers and executives of tomorrow with leading-edge developments in artificial intelligence, prediction and data visualization, combined with a strong business foundation.

Running Machine Learning Systems in Production


Machine learning engineering is the practice of applying machine learning science to production systems. It requires expertise in both machine learning methods and software engineering. In practice, few individuals have sufficiently deep experience in both fields to act as sole practitioners. Scientists and engineers instead must work together, leveraging the skill and experience of one another, to build state-of-the-art machine learning enabled systems. In this masterclass, Garrett Smith, founder of Chicago ML and creator of Guild AI, teaches the fundamentals of machine learning engineering.

Medopad raises $25M led by Bayer to develop biomarkers tracked via apps and wearables – TechCrunch


Medopad, the UK startup that has been working with Tencent to develop AI-based methods for building and tracking "digital" biomarkers -- measurable indicators of the progression of illnesses and diseases that are picked up not with blood samples or in-doctor visits but using apps and wearables, has announced another round of funding to expand the scope of its developments. It has picked up $25 million led by pharmaceuticals giant Bayer, which will be working together with Medopad to build digital biomarkers and therapeutics related to heart health. Medopad said it is also working on separate biomarkers related to Parkinson's, Alzheimer's and Diabetes. The Series B is being made at a post-money valuation of between $200 million and $300 million. In addition to Bayer, Hong Kong firm NWS Holdings and Chicago VC Healthbox also participated.

Kernel Stein Tests for Multiple Model Comparison Machine Learning

We address the problem of non-parametric multiple model comparison: given $l$ candidate models, decide whether each candidate is as good as the best one(s) or worse than it. We propose two statistical tests, each controlling a different notion of decision errors. The first test, building on the post selection inference framework, provably controls the number of best models that are wrongly declared worse (false positive rate). The second test is based on multiple correction, and controls the proportion of the models declared worse but are in fact as good as the best (false discovery rate). We prove that under appropriate conditions the first test can yield a higher true positive rate than the second. Experimental results on toy and real (CelebA, Chicago Crime data) problems show that the two tests have high true positive rates with well-controlled error rates. By contrast, the naive approach of choosing the model with the lowest score without correction leads to more false positives.

Combining No-regret and Q-learning Artificial Intelligence

Combining No-regret and Q-learning Ian A. Kash University of Illinois, Chicago, IL Michael Sullins University of Illinois, Chicago, IL Katja Hofmann Microsoft Research, Cambridge, UK Abstract Counterfactual Regret Minimization (CFR) has found success in settings like poker which have both terminal states and perfect recall. We seek to understand how to relax these requirements. As a first step, we introduce a simple algorithm, local no-regret learning (LONR), which uses a Q-learning-like update rule to allow learning without terminal states or perfect recall. We prove its convergence for the basic case of MDPs (and limited extensions of them) and present empirical results showing that it achieves last iterate convergence in a number of settings, most notably NoSDE games, a class of Markov games specifically designed to be challenging to learn where no prior algorithm is known to achieve convergence to a stationary equilibrium even on average. 1 Introduction V ersions of counterfactual regret minimization (CFR) [50] have found success in playing poker at human expert level [10, 41] as well as fully solving nontrivial versions of it [8]. CFR more generally can solve extensive form games of incomplete information. It works by using a no-regret algorithm to select actions. In particular, one copy of such an algorithm is used at each information set, which corresponds to the full history of play observed by a single agent. The resulting algorithm satisfies a global no-regret guarantee, so at least in two-player zero-sum games is guaranteed to converge to an optimal strategy through sufficient self-play. However, CFR does have limitations. It makes two strong assumptions which are natural for games such as poker, but limit applicability to further settings. First, it assumes that the agent has perfect recall, which in a more general context means that the state representation captures the full history of states visited (and so imposes a tree structure). Current RL domains may rarely repeat states due to their large state spaces, but they certainly do not encode the full history of states and actions. Second, it assumes that a terminal state is eventually reached and performs updates only after this occurs.

Partners Launch Artificial Intelligence Healthcare Guidance Platform


Carmel-based MJ Insurance, one of the nation's largest privately-held insurance agencies, has partnered with Chicago-based HealthJoy, a healthcare guidance platform that helps employees make informed healthcare decisions and provides a variety of cost containment strategies to help lower employer healthcare costs. Together, the partners will be offering an artificial intelligence guidance system to healthcare coverage seekers. HealthJoy advances current telemedicine offerings by providing an artificial intelligence virtual assistant along with a team of live doctors and healthcare professionals to help employees navigate the complexities of the healthcare system and decisions surrounding their employer-sponsored health plans. Unlike other health engagement apps, HealthJoy offers a direct line of communication between the employer and the user by connecting directly to the employee's smartphone device. HealthJoy integrates with any employer benefit plan and has shown a positive ROI for companies within 90 days.