Dallas 2018 Schedule

TIME	DESCRIPTION	ROOM	TOPIC/SPEAKER
09:00-10:00	REGISTRATION	SLC Entrance
10:00-10:10	OPENING PLENARY	SLC102	Int’l Data Engineering and Science Association
10:10-10:40	SESSION	SLC102	Practical and Collaborative Method to Jump Start into Machine Learning with Jupyter Notebooks and Google Collab by Tarek Hoteit, Thomson Reuters
10:10-10:40	SESSION	SLC303	Un-Siloing Data Science Team by Aravind Chiruvelli, ThoughtWorks
10:45-11:15	SESSION	SLC102	Knowledge Engineering: How Artificial Intelligence Has Transformed from Magic to Method through Applied Science by Charlie Burgoyne, Valkyrie Intelligence
10:45-11:15	SESSION	SLC303	Process Mining Meets AI and ML by Viswanath Puttagunta, DIVERGENCE.ai
11:20-11:50	SESSION	SLC102	Performing Data Analytics that Scales by Sergey Maydanov, Intel Corporation
11:20-11:50	SESSION	SLC303	Getting Plugged Into Data Science by Caitlin Hudon, Web.com
11:55-12:25	SESSION	SLC102	Predictive Analytics: Digital Market Research & Challenges by Dr. Kannappan Ramu, Saihasys Inc
11:55-12:25	SESSION	SLC303	Is AI Hype or Will it Transform Your Business? by Angela Hood, ThisWay Global
12:30-13:30	LUNCH BREAK
13:30-14:30	PANEL	SLC102	Discussion of Latest Big Data Technologies and How the Technologies Can Be Applied to Certain Use Cases to Solve Big Data Problems by Jerry Watson, Enterprise Metrics / Sadu Hegde, CorData Solutions / Padmanand Warrier, Independent Investor and Board Advisor
14:35-15:05	SESSION	SLC102	Leveraging Blockchain to Transform the Digital Supply Chain by Vipul Tiwari, Reni Analytics
14:35-15:05	SESSION	SLC303	Digital Cracks in Banking and the Rising Impact of FinTechs by Sidharth Nandi, Digital TaaS
15:10-15:40	SESSION	SLC102	Combining Machine Learning and Blockchain To Create Greater Trust by Mark Lynd, Relevant Track
15:10-15:40	SESSION	SLC303	Inspire and Engage Data Science Novices by Reducing the R Learning Curve by Jesse Mostipak, Teaching Trust
15:45-16:15	SESSION	SLC102
15:45-16:15	SESSION	SLC303	Using Machine Learning to Diagnose Patients with Temporomandibular Disorders (TMD) by Robert Chong, Clockwork Solutions
16:20-16:50	SESSION	SLC102
16:20-16:50	SESSION	SLC303	Catching Medication Adverse Events Before They Occur by Boryana Manz, PCCI
18:00-20:30	Networking Party	British Beverage Company

Topic Description:

Tarek Hoteit

Presentation Topic: Practical and collaborative method to jump-start into machine learning with Jupyter Notebooks and Google Collab.

Brief Description:

Not at all machine learning enthusiasts are alike, and, hence, setting up code environment or training the data model can many times be overwhelming for newcomers in the field of machine learning and deep learning. Data scientists may not have the necessary skills in setting up development environments, and, programmers, may not necessarily have the data scientists skills for preparing data sets and evaluating machine learning models. Furthermore, data scientists and programmers together in some enterprise may lack the collaboration platforms to work together on such projects. Even though most of the books on machine learning using some programming language X provide readers with instructions for setting up the coding environment, such chapters can derail the process of getting started if one gets stuck in the setup. Alternatively, collaborative platforms using Jupyter Ipython Notebooks (http://jupyter.org) and Google Collab (https://colab.research.google.com) provide a quick starter for programmers and data scientists alike to develop and collaborate on machine learning algorithms through open source without getting stuck with the nuances of setting up their environments or having to depend on commercial products. The talk revisits the value of Jupyter notebooks for newcomers to the field of AI by showcasing live examples and sharing sources of machine learning algorithms running online using Jupyter notebooks and providing a set of guidelines for implementing such technology in the workplace or leveraging existing ones in the market such as Google Collab. The talk is intended to encourage machine learning enthusiasts to enter the field through a practical method that minimizes the stress from the overwhelming material available on the Internet on how to get started with machine learning.

Aravind Chiruvelli

Presentation Topic: Un-siloing Data Science Teams

Brief Description:

As the notion of Big Data is moving past the hype and buzz, many groups and companies are still struggling to reap Big Value from the big data. Primarily the process of value creation is a multidimensional and technology dimension is the only one with tremendous focus. The other dimensions: people(teams), culture and value proposition are often overlooked. While agile methodologies have proven to be very effective for rapid delivery cycles, there are some challenges in adopting the same approaches when it comes Data Science projects. While Data Science inherently follows test&learn cycles and is effective for stand-alone data science teams, often poses challenges when working within the application development team. This talk focuses on some gaps and challenges in embedding a data science team within traditional application team and continuously derive value from the innovative solutions from data science.

Charlie Burgoyne

Presentation Topic: Knowledge Engineering: How Artificial Intelligence has transformed from magic to method through applied science.

Brief Description:

While AI hype abounds, there are incredible examples of the dramatic impact AI and machine learning can have on industry. We’ll discuss a few little-known case studies that demonstrate the tangible value AI can provide and the thread that connects them all. That thread is the scientific field of “knowledge engineering” where the structure and relationships of data are optimized for business insights.

Viswanath Puttagunta

Presentation Topic: Process Mining meets AI and ML

Brief Description:

This talk introduces you to Process Mining, a relatively new field that allows for analysis of business/other processes based on event logs. Process Mining automatically discovers what people are really doing (instead of what they are supposed to be doing), identify deviations, bottlenecks in processes to name a few. This also forms the foundation to augment traditional Machine Learning (Data Mining) to the discovered process. This allows for example to predict the probability that a claim may get rejected much earlier in the process.

Sergey Maydanov

Presentation Topic: Performing data analytics that scales

Brief Description:

This talk will cover challenges of performing data analytics at scale. The tools which are often used for prototyping are not designed to scale to large problems. As a result organizations have to have a dedicated team that takes a prototype created by a data scientist and deploys it in a production environment. In certain businesses where the need of frequent model changes arises this is simply unsustainable.
A new approach is required for addressing both scalability and productivity aspects of a data science that combines two distinct worlds, the best of HPC world and the best of database worlds.
Starting with a brief overview of scalability aspects with respect to modern hardware architecture we will characterize what big data problem is, its inherit characteristics and how these map onto a modern data analytics software design choices.
We will also overview some data analytics use cases illustrating that the Big Data is not tied to cluster of machines within a data center.
Finally, we will present a few case studies using industry known data analytics libraries, such as Scikit-learn, Intel® Data Analytics Acceleration Library and others, to illustrate the scalability aspect.

Caitlin Hudon

Presentation Topic: Getting Plugged Into Data Science

Brief Description:

Field notes from a data scientist on getting plugged into the data science world, including topics like:

+ How to network
+ Career tracks within “data science”
+ Advice for finding your first data science job
+ The data science skills they don’t teach in school
+ A secret for overcoming imposter syndrome

Dr. Kannappan Ramu

Presentation Topic: Predictive Analytics: Digital Market Research & Challenges

Brief Description:

Today’s companies want to decrease data storage on their local sites and improve the time-to-value for their analytical data. On-premises big data analytics have significant operational limitations and companies desire to transform the way their real-time data is ingested, processed and delivered to provide meaningful business predictions and facilitate accurate decision making, while also lowering cost. Now and in the future to uncover niche audience/customer needs has become increasingly challenging for market researchers due to the wealth of data and emerging technologies that are readily available.

In particular, this talk will feature of recent work on Sales Forecasting, customer purchase pattern/behavior, Fraud Investigations and digital Marketing analytics.

Angela Hood

Presentation Topic: Is AI Hype or Will it Transform Your Business?

Brief Description:

Mrs. Hood will first help define the important difference between AI and IA. She will then define a three step process in evaluating proposed AI products for your business. The presentation will highlight how people use AI in everyday life and where it can best be applied in business. Finally, she will lay actionable steps that each business can and should take to be sure their business is not left behind.

Jerry Watson, Sadu Hegde, Padmannand Warrier

Panel Topic: Discussion of latest big data technologies and how the technologies can be applied to certain use cases to solve big data problems

Brief Description:

Jerry Watson will spend the first 15-20 minutes presenting slides of the primary big data technologies in use today, such as Hadoop (Pig, Hive, etc.), Elastic Map Reduce, Apache Spark, Amazon Web Services, IBM Watson Analytics, Microsoft Cortana/Azure, Google Analytics. The next 15-20 minutes will be Jerry asking the 2 panel members questions about matching the technologies to use cases to solve big data problems.

Vipul Tiwari

Presentation Topic: Leveraging Blockchain to transform the Digital Supply Chain

Brief Description:

– Future of Blockchain
– Origins of Blockchain
– Current Usage
– Applications & Use Cases of BlockChain in various industries
– Looking Forward

Sidharth Nandi

Presentation Topic: Digital cracks in banking and the rising impact of FinTechs

Brief Description:

Banking models which have stayed resilient for centuries rooted in their traditional ways of serving customers have started to experience cracks driven by the exposure of their inherent flaws and fintech powered digital disruption with the dawn of 4th industrial revolution. The rise of digital-only banks has started to fundamentally question the premise of a traditional brick-n-mortar outfit. The incumbent financial institutions and fintechs have started to identify ways towards a consumer centric view to reposition their relationship from competition to collaboration. What are the traits of survivors and disruptors? What does it mean to the consumers?

Mark Lynd

Presentation Topic: Combining Machine Learning and Blockchain To Create Greater Trust

Brief Description:

The natural intersection of machine learning and blockchain is greater trust. Organizations that are successful in creating more trust through this intersection by building smart immutable applications will ultimately win more customers and build longer more meaningful relationships with these customers. Here are ways to examine the value of this combination of blockchain and machine learning:

1. Creating intelligent data-driven experiences

2. Examples of smart, immutable, tamper-proof applications

3. Developing offerings with trust as their notable characteristic

2018 will be a huge year for machine learning and blockchain, but it is the ability to combine them that drives differentiation and provide greater competitive advantage.

Jesse Mostipak

Presentation Topic: Inspire and engage data science novices by reducing the R learning curve

Brief Description:

One of the biggest pain points for both teachers and learners of data science in R is navigating the often unspoken prerequisite skills and content knowledge necessary to successfully apply R to data science problems. In this talk, R for data science educators will learn actionable strategies to more effectively bring learners up to speed, while learners will develop strategies to identify and address their own knowledge gaps.

By incorporating learnings from the establishment of a data-driven culture at Teaching Trust, coupled with her experience creating and leading the R for Data Science Online Learning Community, Jesse will share strategies that can be immediately implemented with groups of any size in order to more quickly develop data science skills in R. These include methods for identifying individual gaps in knowledge, establishing mentor and learner relationships, incorporating best practices from the field of education, and bolstering fundamental computer science skills.

Dr. Valarie J. Bell

Presentation Topic: Data’s ‘Social DNA’: Decoding the People Behind the Data

Brief Description:

Data science generally aims at one very large target: quantifying & qualifying human behavior or a dimension thereof. The so-called ‘data nerds’ on one riverbank oppose the ‘warm and fuzzy’ social scientists on the other. A fast-moving river separates them; try crossing it and risk drowning in currents of seemingly incompatible theoretical & methodological paradigms. “Data nerds don’t understand people & lack true insights, ” or “Social scientists lack data skills & can’t code,” has been preached before. If that’s true, how can we find the ideal, middle ground? I will present specific steps to help professionals identify the true ‘social DNA’ (thoughts, behaviors, feelings, emotions, motives and motivations) embedded in data. I will demonstrate through real-world examples why predictions ACTUALLY prove wrong, why promotions, product launches, brands and campaigns REALLY fail, why people never seem to do what you want or expect them to do, & how to truly understand the very different subcultures & demographic population segments critical to your organization’s maximizing all that wonderful data you’ve spent so much time, effort, and money to collect.

Robert Chong

Presentation Topic: Using Machine Learning to Diagnose Patients with Temporomandibular Disorders (TMD)

Brief Description:

In this presentation, Mr. Chong will outline a methodology in which machine learning algorithms can be applied to patient sensor data in order to quantify the degree in which they suffer TMD. This presentation will give a brief introduction to TMD and discuss why TMD is such a challenging problem to solve. He will then demonstrate how Electromyography (EMG) sensor data can be decomposed to generate salient features. These features are then fed into dynamic models to generate a physical representation of the system. He will conclude the presentation by discussing how one may implement an anomaly detector that is personalized for a patient.

Boryana Manz

Presentation Topic: Catching medication adverse events before they occur

Brief Description:

Within the last 10 years, healthcare providers have transitioned to electronic health records, enabling the development and implementation of data-driven approaches at scale to improve patient care and hospital operations. The challenges are balancing complex outcomes and integrating the solutions within the provider workflow. Over the last five years, PCCI (Parkland Center for Clinical Innovation) has pioneered the use of advanced analytics and artificial intelligence to solve healthcare’s challenges at Parkland and beyond.

About 5% of hospitalized patients in the United States experience an adverse drug event (ADE) – any harm experienced that is related to medication. ADEs especially occur during care transitions such as hospital admissions and are influenced by a complex etiology of patient, medication, provider, and socioeconomic-specific factors. Proactive identification of patients at high risk and timely interventions by pharmacists, can significantly reduce preventable ADEs.
PARADE (Patients at Risk for ADEs) is a multifaceted predictive score developed for Parkland Hospital, Dallas, to stratify newly admitted patients. It generates real-time actionable worklists within the electronic health record that trigger timely interventions by care teams. The PARADE model reconciles multivariable logistic regression models for best practices and ADEs. Post-implementation data reveals that consults for high-risk patients have tripled without additional pharmacy resources. PARADE-enhanced workflow has enabled efficient use of limited resources for improved outcomes.