Research
I work on both theoretical and application-based topics in optimization and data analysis. Application areas include pharmaceutical supply chain, clinical trial operations, and inventory management. My research can be generally summarized in the following streams:
Clinical trial operations
The process of clinical trials for new medical treatments is a challenging undertaking. The development of the treatments themselves is an expensive and time-consuming task. Beyond that is the more subtle issue of evaluating the effectiveness of any such treatment. Poorly designed management of those trials is not just expensive, it can lead to the collection of useless data. My approach to this problem entails the establishment of optimization and simulation models that can improve the efficiency of clinical trials. Those models speak to optimal selection of patient trial sites along with the requisite supply chain required to provide treatment.
Research papers in this area:
(*: undergraduate students)
Ninh, A., Bao, Y., Mcgibney, D. and Nguyen, T., 2024. Clinical trial site selection with probabilistic constraints. European Journal of Operations Research, https://doi.org/10.1016/j.ejor.2024.03.013.
Rubio-Herrero, J., Ninh, A. and Lefew, M., 2023. Improving the performance of supply chains in clinical trials with delays: an optimization approach to determining the number of recruitment sites. Annals of Operations Research, 1-21.
Ninh, A., Melamed, B. and Zhao, Y., 2020. Analysis and optimization of recruitment stocking problems. Annals of Operations Research, 295(2), 747-767.
Lefew, M., Ninh, A. and Anisimov, V., 2020. End-to-end drug supply management in multi-center trials. Methodology and Computing in Applied Probability, pp.1-15.
Ninh, A., LeFew, M. and Anisimov, V., 2019, December. Clinical trial simulation: Modeling and practical considerations. In 2019 winter simulation conference (WSC) (118-132). IEEE.
Fleischhacker, A., Ninh, A. and Zhao, Y., 2015. Positioning inventory in clinical trial supply chains. Production and Operations Management, 24(6), 991-1011.
Magazine:
Clinical Trial Insight, December 2020: "Moving targets".
Machine learning applications
Recent years has seen a fast-growing trend of data analytics in operations and supply chain management. Increasing availability of data and technological advancement in machine learning create new opportunities to further improve how firms operate. My research in this direction focuses on applying machine learning techniques to address important decision problems from various application contexts such as pharmaceutical supply chains, or intelligent transportation systems.
Research papers in this area:
(*: undergraduate students)
Zhu, X., Ninh, A., Zhao, H. and Liu, Z.M., 2021. Demand forecasting with supply‐chain information and machine learning: Evidence in the pharmaceutical industry. Productions and Operations Management, 30(9), 3231-3252.
Pham, M., Ninh, A., Le, H. and Liu, Y., 2020. An efficient algorithm for minimizing multi non-smooth component functions. Journal of Computational and Graphical Statistics,1-9.
*Morris, J. and Ninh, A., 2023. Learning and Planning for Self-Driving Ride-Hailing Fleets. Submitted.
Optimization
Log-concavity plays an important role in economics, operations management and particularly optimization since it helps ensure tractability of the associated decision problem. For instance, in classic revenue management models, log-concavity of the underlying demand distributions guarantees the concavity/unimodality of the objective functions (Ninh et al. 2020). I am also interested in the moment bounding problems when the random variables are discrete and subject to log-concave/log-convex constraints.
Research papers in this area:
(*: undergraduate students)
Ninh, A., 2021. Robust newsvendor problems with compound Poisson demands. Annals of Operations Research, 302(1), 327-338.
Ninh, A., Shen, Z.J.M. and Lariviere, M.A., 2020. Concavity and Unimodality of Expected Revenue Under Discrete Willingness to Pay Distributions. Production and Operations Management, 29(3), 788-796.
Ninh, A., Hu, H. and Allen, D*., 2019. Robust newsvendor problems: Effect of discrete demands. Annals of Operations Research, 275(2), 607-621.
Prékopa, A., Ninh, A. and Alexe, G., 2016. On the relationship between the discrete and continuous bounding moment problems and their numerical solutions. Annals of Operations Research, 238(1-2), 521-575.
Ninh, A. and Prékopa, A., 2013. Log-concavity of compound distributions with applications in stochastic optimization. Discrete Applied Mathematics, 161(18), 3017-3027.