Spring 2022 Newsletter
New LRI Associate Director
Cesar Bandera is our new LRI Associate Director and will serve as Acting Director for the 2022-2023 academic year.
Cesar Bandera's research interests follow three paths: cross-cultural experiential entrepreneurship education, innovation-driven business incubation, and mobile healthcare innovation. In addition to teaching most NJIT courses in entrepreneurship, he is also a serial entrepreneur and currently the founding partner of an m-health company that conducts wireless public health campaigns for CDC, NIH, EPA, and foreign ministries of health. Bandera help students and alumni launch their own startups and secure funding; if their product or service is in his technology wheelhouse, helps with product development as well.
Faculty Seed Grants
FY 2022 Faculty Seed Grants Winners
Principal Investigator: Joseph Micale
Department: MTSM
Project Title: Textual Analysis & Machine Learning in Sustainability Reporting
Co-Principal Investigator(s): Mingying Cheng (Fordham University)
Principal Investigator: Jinghua (Carolyn) Wang
Department: MTSM
Project Title: An Investigation of the Predictability of Uncertainty on Bitcoin Returns
Principal Investigator: Aichih (Jasmine) Chang
Department: MTSM
Project Title: Plastic Recycling Supply Chain Driven by Data Science and Blockchain Technology
Co-Principal Investigator(s): Jim Shi (MTSM)
FY 2021 Faculty Seed Grants Semi-annual Update
Exploring unicorn ventures in Space exploration: Raja Roy, Shanthi Gopalakrishnan, Xi Zhang
Save the Date
Leir Research Institute Virtual Annual Conference
Thursday, September 1st, 2022
Disruptive Technologies, Regulations, Business -
Implications in the Real Estate and Property Tech Industry
Visit our conference website for updates.
Research Spotlights
Xinyuan Tao
Economic Policy Uncertainty and the Cross-Section of Corporate Bond Returns
This paper finds that economic policy uncertainty (EPU) is a systematic risk factor priced in the cross-section of corporate bonds. Bonds with high EPU beta have low expected returns. As bonds with high sensitivities to policy uncertainty provide good hedge and draw high demand from investors, thus leading to higher prices and lower required rates of returns for these bonds. This negative premium is robust to controlling for conventional risk factors, bond characteristics, macroeconomic conditions and more importantly, beyond the general economic uncertainty. The effect of policy risk is pervasive across different firms and over time. It is more severe when the economy has higher uncertainty, and stronger for firms exposed to higher uncertainty. The EPU risk effect is greater for firms with higher earnings exposure to policy uncertainty, dependence on external financing and effective tax rates, lower pre-tax interest coverage, and operating in regulation-intensive industries. In addition, there is a clear evidence of spillovers and variations in the effects of economic policy uncertainty across countries.
Tao, X., B. Wang, J. Wang, and C. Wu. 2022. “Economic Policy Uncertainty and the Cross-Section of Corporate Bond Returns.” Journal of Fixed Income, forthcoming.
Are missing values important for earnings forecasts? A machine learning perspective
Analysts' forecasts are one of the most common and important estimators for firms' future earnings. However, they are challenging to fully utilize because of missing values. This study applies machine learning techniques to estimate missing values in individual analysts' forecasts and subsequently to predict firms' future earnings based on both estimated and observed forecasts. After estimating missing values, forecast error is reduced by 41% compared to the mean forecast, suggesting that missing values after estimating are indeed useful for earnings forecasts. We analyze multiple estimation methods and show that the out-performance of matrix factorization (MF) is consistent using different evaluation measures and across firms. Finally, we propose a stochastic gradient descent based coupled matrix factorization (CMF) to augment the estimation quality of missing values with multiple datasets. CMF further reduces the error of earnings forecasts by 19% compared to MF with a single dataset.
Uddin, A., X. Tao, C. C. Chou, and D. Yu. 2022. “Are missing values important for earnings forecasts? A machine learning perspective.” Quantitative Finance, forthcoming.
Yi Chen
The study of online publishing ecosystem with computational advertising and user privacy protection
The online publishing ecosystem employs big data and computational models for revenue generation and readership attraction. The goal of our research is to analyze user behaviors, publishers’ strategies, advertisers’ reactions using machine learning techniques to understand the behavioral, economic and regulatory issues in the online publishing ecosystem, and to improve user experience, protect user privacy, benefit economy and contribute to a healthy ecosystem. This project is in collaboration with Hearst and Forbes media.
Zhao, S., Bharati, R., Borcea, C. M., Chen, Y. (2020). Privacy-Aware Federated Learning for Page Recommendation. 2020 IEEE International Conference on Big Data (IEEE BigData).
Zhao, S., Kalra, A., Borcea, C. M., Chen, Y. (2020). To be Tough or Soft: Measuring the Impact of Counter-Ad-blocking Strategies on User Engagement. Proceedings of The Web Conference (WWW).
Wang, C., Chen, Y. (2020). Topical classification of domain names based on subword embeddings. Electronic Commerce Research and Applications, 40, 37.
Wang, C., Zhao, S., Kalra, A., Borcea, C. M., Chen, Y. (2019). Webpage Depth Viewability Prediction using Deep Sequential Neural Networks. IEEE Transactions on Knowledge and Data Engineering (TKDE), 31(3).
Kalra, A., Wang, C., Borcea, C. M., Chen, Y. (2019). In ACM (Ed.), Reserve Price Failure Rate Prediction with Header Bidding in Display Advertising (pp. 2819-2827). The 25th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD).
Zhao, S., Kalra, A., Wang, C., Borcea, C. M., Chen, Y. (2019). In IEEE (Ed.), Ad Blocking Whitelist Prediction for Online Publishers. The 2019 IEEE International Conference on Big Data (IEEE Big Data).
Understanding and assisting patient decision making using machine learning and natural language processing techniques.
Studies show that patient engagement drives better health outcomes. The long-term goal of this research is to understand patient needs, concerns, and decision making processes and further to develop appropriate patient engagement and intervention strategies and to improve healthcare outcomes by developing cutting-edge technologies in data science. We are developing machine learning and natural language processing techniques that allow us to better understand the factors and influences in patient decision making as expressed in online health forums, such as PatientsLike Me, MedHelp, and Cancer Survivor Network.This project is funded by the Leir Foundation.
Li, M., Shi, J., Chen, Y., (2022). Identifying Influences in Patient Decision-Making Processes in Online Health Communities. Journal of Medical Internet Research, accepted.
Li, M., Gao, W., Chen, Y. (2020). A Topic and Concept Integrated Model for Thread Recommendation in Online Health Communities. 29th ACM International Conference On Information and Knowledge Management (CIKM).
Shi, J., Chen, Y. (2020). User and Context Integrated Experience Mining in Online Health Communities. 29th ACM International Conference On Information and Knowledge Management (CIKM).
Liu, Y., Shi, J., Chen, Y. (2019). Thread Structure Learning on Online Health Forums with Partially Labeled Data. IEEE Transactions on Computational Social Systems (TCSS), 6(6).
Li, M., Shi, J., Chen, Y. (2019). In IEEE (Ed.), Analyzing Patient Decision Making in Online Health Communities (pp. 177-184). IEEE International Conference on Healthcare Informatics (ICHI).
Knowledge discovery for clinic decision support using machine learning and natural language processing.
The wide prevalence of Electronic Health Record (EHR) systems provides a large and fast-growing volume of clinic data. Yet the potential of what such big data can bring to support clinic decision making is still not fully exploited. We are working on developing and applying the state-of-the art machine learning and natural language processing techniques to identify and extract key information from unstructured clinic notes for downstream machine learning models and to make risk assessment for clinic decision making that provide effective, and at the same time, safe treatment to patients. This project is funded by NIH and the Leir Foundation.
Shi, J., Gao, X., Kinsman, W. C., Ha, C., Gao, G. G., Chen, Y. (2022). DI++: A deep learning system for patient condition identification in clinical notes. Artificial Intelligence in Medicine, 123.
Shi, J., Gao, X., Ha, C., Wang, Y., Gao, G. , Chen, Y. (2020). Patient ADE Risk Prediction through Hierarchical Time-Aware Neural Network Using Claim Codes. 2020 IEEE International Conference on Big Data (IEEE BigData).
Traffic and user location prediction for better service offering using machine learning.
Network traffic and user location prediction on smart phones are useful to improve wireless network service quality, enhance system performance and enable location-based mobile services. We are developing machine learning techniques for accurate traffic and user location prediction. At the same time, using federated leaning, our system also protects user privacy and reduces network bandwidth consumption. This project is in collaboration with and funded by AT&T research.
Jiang, X., Zhao, S., Jacobson, G., Jana, R., Hsu, W.-L., Talasila, M., Aftab, S. A., Chen, Y., Borcea, C. M. (2021). In IEEE (Ed.), Federated Meta-Location Learning for Fine-Grained Location Prediction. IEEE International Conference on Big Data (IEEE BigData).
Zhao, S., Jiang, X., Jacobson, G., Jana, R., Hsu, W.-L., Rustamov, R., Talasila, M., Aftab, S. A., Chen, Y., Borcea, C. M. (2020). Cellular Network Traffic Prediction Incorporating Handover: A Graph Convolutional Approach. The 17th Annual IEEE International Conference on Sensing, Communication and Networking (SECON).