Paul Russo, Ph.D. Vice Provost and Dean

Recent research examines computational methods in social networks and mobile applications by developing and empirically evaluating causal models. Studies focus on social sharing in online communities that are influenced by user motivations, trust, and network properties. This computational social science on the web promises to improve AI algorithms on platforms like Facebook, Instagram, Google Maps, and Tinder. In earlier projects, such as those at Texas Instruments, used Pattern Analysis and Machine Intelligence methods—e.g., clustering and Horn&Schunk algorithm—for image and streaming video analysis for machine vision and tracking moving objects. Other projects have bridged engineering and physics methods to build data collection systems that identify particles emitted as deep space materials degrade. Ethnographic studies in the use of technology in distributed science collaborations. Active peer reviewer. Co-PI for a $15M award to create the CUNY Center for Big Data.

Teaches graduate courses in Social Computing, Digital Media, Information Systems, Technology Entrepreneurship, Organizational Behavior, and thesis supervision.

Andy Catlin, Program Director

Data scientist and data system developer with expertise in financial instrument pricing and forecasting using weighted Black Scholes Merton model as well as incorporating yield curve analysis into emerging markets products. Recent projects focus on real time market feeds and server-based cascading triggers as well as migrating client-server systems into web and cloud hosted solutions. Previous projects focused on incorporating artificial intelligence and neural networks in specialized applications; C++ financial libraries; security authentication bottleneck correction; patient-days forecasting model using Box Jenkins; production replication architectures; and multi-phase commit and log-shipping architectures. Founder of multiple tech startups, including the Hudson Technology Group (acquired by Incepta), which served major clients including Fidelity Investments; Smart Money; Donaldson, Lufkin and Jenrette (DLJ); Manufacturers Hanover Trust; National Football League; and The Wall Street Journal.

Teaches graduate courses in Analytics Programming, Recommendation Systems, Regression Modeling, Network Analysis, Natural Language Processing, and Neural Networks.

Sergey Fogelson, Instructor

Data scientist and data product architect with expertise in media and advertising-related pricing, scheduling, and forecasting, and in large-scale anonymized identity models. Major projects have included building petabyte-scale data warehouses for media asset management and consumption analysis use-cases; fault-tolerant data products utilizing human-in-the-loop machine learning algorithms for back-office financial applications; and risk-scoring algorithms for 3rd-party cybersecurity vendor risk management. Doctoral research examined hierarchical category learning mechanisms in the visual system utilizing supervised learning algorithms applied to functional magnetic resonance imaging (fMRI) data.

Teaches graduate courses in Computational Statistics, Linear Algebra, and Machine Learning.

Lawrence Fulton, Instructor

Health data scientist with expertise in machine learning for image recognition, especially applied to 4D MRIs, and in healthcare simulation in Java, PySim, RSimmer, and ProModel.  Current research focuses on the use of mathematical programming for improving health system performance in the areas of cost, quality, and access. Previous work focused on the application of artificial intelligence and neural networks in specialized applications, the use of Python and R applications to health data science problems, and the application of hierarchical forecasting models using both imagery and time series data simultaneously. Published in over 70 peer-reviewed journals.

Teaches graduate courses in Data Analytics, Machine Learning, and Structured Data Management.

Jeff Nieman, Instructor

Data scientist and project manager with expertise in visualization and predictive modeling for Fortune 500 companies, including Ford and Cisco. Current work involves building systems engineering and model governance and support approaches using open source and COTS software like Alteryx, Qlik, Tableau and DataRobot. Led initiative to leverage data science optimization across complex server ecosystems, increasing server performance and minimizing risk through automated system and usage monitoring. Other key projects have included automating quoting for service renewals, tracking quoting and booking against opportunity data, enabling users to embed snapshots of visualizations in email notifications, and developing a major app for all Ford drivers. 

Teaches graduate courses in Predictive Modeling, Mathematics, Statistics, Machine Learning, and Project Management.

David Sweet, Instructor

David Sweet is a quantitative trader and former machine learning engineer at 3Red Partners, a social media company in New York. He co-founded a cryptocurrency trading company, Bedford Ave. Trading/Galaxy Digital Trading, which was acquired by the first crypto-focused investment bank. He was the principal author of KDE 2.0 Development, published by Macmillan in 2000, and was a contributing author to Special Edition: Using KDE, also published by Macmillan, both books about computer programming and experimental methods for ML engineers working in finance and technology. His book, Tuning Up: From A/B Testing to Bayesian Optimization, which is an extension of his lectures on tuning quantitative trading systems over the past three years, will be published by Manning this year. In addition, he has published research in Nature, Physical Review Letters and Physical Review D, and helped fashion the Nature research into an exhibit at the Museum of Mathematics in New York City. David holds a Ph.D. in physics from the University of Maryland, and a B.S. in physics and B.A. in mathematics from Duke University.

Teaches Predictive Analytics.