Mohankumar, N. M., Hefley, T. J., & Boyle, W. A. (2023). Robustness of point process species distribution models to misspecified temporal support. (Under review in Journal of Agricultural, Biological and Environmental Statistics)
Mohankumar, N. M., Jain, M., Wan, H., Ganguli, S., Anderson, D., & Wilson, K. (2023). Principal Component Regression to Study the Impact of Economic Factors on Disadvantaged Communities (Under review in Computers, Environment and Urban Systems)
Wan, H., Ganguli, S., Jain, M., Anderson, D., Mohankumar, N. M., & Wilson, K. (2023). Areal interpolation of population projections consistent with different SSPs from 1-km resolution to block level based on USA Structures dataset. Computers, Environment and Urban Systems, 105, 102024. https://www.sciencedirect.com/science/article/pii/S019897152300087X Mohankumar, N. M., Hefley, T. J., Silber, K. M., & Boyle, W. A. (2023). Data fusion of distance sampling and capture-recapture data. Spatial Statistics, 55, 100756. https://www.sciencedirect.com/science/article/pii/S2211675323000313 (Presented in the North American Ornithological Conference 2020 (virtual), Joint Statistical Meetings 2019 (virtual), Three minute thesis competition 2020 (Kansas State University)
Jain, M., Mohankumar, N. M., Wan, H., Ganguly, S., Wilson, K. D., & Anderson, D. M. (2023). Training Machine Learning Models to Characterize Temporal Evolution of Disadvantaged Communities. arXiv preprint arXiv:2303.03677. https://arxiv.org/abs/2303.03677
Mohankumar, N. M., & Hefley, T. J. (2022). Using machine learning to model nontraditional spatial dependence in occupancy data. Ecology, 103(2), e03563. https://esajournals.onlinelibrary.wiley.com/doi/10.1002/ecy.3563 (Presented in the International Conference on Environmental and Medical Statistics 2020, Joint Statistical Meetings 2019 (Denver. CO), Three Minute Thesis Competition 2019 (Kansas State University), Graduate Research Forum 2019 (Kansas State University), SciComm 2019 (Kansas State University & University of Nebraska-Lincoln.)
Sroor, E.N., W.A. Boyle, H.N. Castro-Miller, K.S. Hobbs, M.D. Reynaldo, K.M. Silber, N.M. Mohankumar, N.A. Wright, and N.E. Freeman. Environmental drivers of shifts in female reproductive investment from quantity to quality in a grassland songbird. (Under review in Methods in Animal Behavior)
Hefley, T. J., Boyle, W. A., & Mohankumar, N. M. (2020). Accounting for location uncertainty in distance sampling data. arXiv preprint arXiv:2005.14316 (Under review in Methods in Ecology and Evolution) https://arxiv.org/abs/2005.14316
Patrignani, A., Mohankumar, N., Redmond, C., Santos, E. A., & Knapp, M. (2020). Optimizing the spatial configuration of mesoscale environmental monitoring networks using a geometric approach. Journal of Atmospheric and Oceanic Technology, 37(5), 943-956. https://doi.org/10.1175/JTECH-D-19-0167.1 (Presented in Capitol Graduate Research Summit 2018, Governor’s Conference on the Future of Water in Kansas 2017 and Graduate Research and the State Poster Session 2017)
The Wildlife Society’s Annual Conference, Spokane, Washington (2022): Fuzzy Boundaries: Machine Learning for Wildlife Ecology symposium Title: Using machine learning to model nontraditional spatial dependence in occupancy data
Predicting the expected number of bio-luminescence sources in the deep sea using Bayesian Additive Regression Trees The main focus in this study is to predict the expected number of bioluminescence sources in each depth in deep sea, using Bayesian regression trees (BART). Data for this analysis consists of number of bioluminescence sources and the depth of the sea it was recorded. BART is mostly used for higher prediction purposes and to determine the distribution of a species where the distribution over a spatial area does not seem continuous. BART is an efficient method to use when we have complex Bayesian hierarchical structures with uncertainty. Metropolis algorithm in MCMC is used to get the posterior distribution of the parameters where the distribution is unknown and the full conditionals are used for other known parameters. BART package in R-Studio was mainly used for this analysis.
Growth rate patterns of tree species in Sinharaja Forest, Sri Lanka (Presented in PGIS research congress 2015, Sri Lanka) This study used spatial point patterns to examine ecological processes associated with the growth rates of tree species in the Sinharaja forest. It is assumed that, growth rate of tree species is related to several factors (e.g. Janzen-Connell and neutral effects, habitat variability, competition, and canopy-gap formation). Main objective of this study is to propose a method that modifies the Ripley's K-function to calculate the growth rates of species in the Sinharaja forest, Sri Lanka and to test whether the growth rate of individual tree species depend on its physical location. Here, we examined growth rates of two canopy species: Mesua nagassarium (Clusiaceae) and Shorea affinis (Dipterocarpaceae) to test this hypothesis. Data on the tree coordinates and diameter at breast height of the trees from 1996 and 2001 censuses from the Sinharaja Forest Dynamic Plot were used to determine their growth rates. The above method coupled with random labeling test was used to test whether the location-specific factors affect the growth rates of tree species.
Determinants of nutritional status and growth rate of children, attending a well baby clinic in Peradeniya, Sri Lanka Objective of this research was to identify the factors effecting the growth rate of children attending a well baby clinic in Sri Lanka. Data were collected from mothers of babies who attend the Baby clinic at Teaching hospital in Peradeniya, Sri Lanka. Multinomial regression, Ordered logit/probit models, Association rule mining and Regression tree methods were used to conduct this analysis.
Factors affecting the special degree selection in Faculty of science, University of Peradeniya, Sri Lanka Data were collected from students in University of Peradeniya, Sri Lanka using a survey to identify the factors that affect the special degree selection in the university. Binomial logistic regression and non-parametric methods were used for this analysis.