About

Murali Haran is a Professor of Statistics and a Graduate Faculty member in Social Data Analytics at Pennsylvania State University. He is affiliated with the departments of Social Data Analytics and Statistics, and his office is located at 302 Pond Laboratory, University Park, PA 16802. His research focuses on social data analytics, applying statistical methods to analyze social data, although specific details of his research interests are not provided on the page. As a faculty member, he contributes to the academic community through teaching and research, and he maintains a professional webpage and a Google Scholar profile for further information about his scholarly work.

Research topics

• Sociology
• Econometrics
• Psychology
• Demography
• Geography
• Medicine
• Mathematics
• Statistics
• Virology
• Surgery
• Biology
• Social psychology
• Microbiology

Selected publications

• Probabilistic Downscaling for Flood Hazard Models

  ArXiv.org · 2025-03-26

  preprintOpen accessSenior author

  Riverine flooding poses significant risks. Developing strategies to manage flood risks requires flood projections with decision-relevant scales and well-characterized uncertainties, often at high spatial resolutions. However, calibrating high-resolution flood models can be computationally prohibitive. To address this challenge, we propose a probabilistic downscaling approach that maps low-resolution model projections onto higher-resolution grids. The existing literature presents two distinct types of downscaling approaches: (1) probabilistic methods, which are versatile and applicable across various physics-based models, and (2) deterministic downscaling methods, specifically tailored for flood hazard models. Both types of downscaling approaches come with their own set of mutually exclusive advantages. Here we introduce a new approach, PDFlood, that combines the advantages of existing probabilistic and flood model-specific downscaling approaches, mainly (1) spatial flooding probabilities and (2) improved accuracy from approximating physical processes. Compared to the state of the art deterministic downscaling approach for flood hazard models, PDFlood allows users to consider previously neglected uncertainties while providing comparable accuracy, thereby better informing the design of risk management strategies. While we develop PDFlood for flood models, the general concepts translate to other applications such as wildfire models.

  PublisherOA PDFDOI

• A class of models for large zero-inflated spatial data

  Journal of Agricultural Biological and Environmental Statistics · 2024-04-29 · 3 citations

  articleOpen accessSenior author

  Abstract Spatially correlated data with an excess of zeros, usually referred to as zero-inflated spatial data, arise in many disciplines. Examples include count data, for instance, abundance (or lack thereof) of animal species and disease counts, as well as semi-continuous data like observed precipitation. Spatial two-part models are a flexible class of models for such data. Fitting two-part models can be computationally expensive for large data due to high-dimensional dependent latent variables, costly matrix operations, and slow mixing Markov chains. We describe a flexible, computationally efficient approach for modeling large zero-inflated spatial data using the projection-based intrinsic conditional autoregression (PICAR) framework. We study our approach, which we call PICAR-Z, through extensive simulation studies and two environmental data sets. Our results suggest that PICAR-Z provides accurate predictions while remaining computationally efficient. An important goal of our work is to allow researchers who are not experts in computation to easily build computationally efficient extensions to zero-inflated spatial models; this also allows for a more thorough exploration of modeling choices in two-part models than was previously possible. We show that PICAR-Z is easy to implement and extend in popular probabilistic programming languages such as and .

  PublisherOA PDFDOI

• Computer Model Calibration Based on Image Warping Metrics: An Application for Sea Ice Deformation

  UNC Libraries · 2024-09-06

  articleOpen access
  PublisherDOI

• Fast Bayesian Inference for Spatial Mean-Parameterized Conway–Maxwell–Poisson Models

  Journal of Computational and Graphical Statistics · 2024-08-21 · 3 citations

  articleOpen accessSenior author

  Count data with complex features arise in many disciplines, including ecology, agriculture, criminology, medicine, and public health. Zero inflation, spatial dependence, and non-equidispersion are common features in count data. There are currently two classes of models that allow for these features-the mode-parameterized Conway-Maxwell-Poisson (COMP) distribution and the generalized Poisson model. However both require the use of either constraints on the parameter space or a parameterization that leads to challenges in interpretability. We propose spatial mean-parameterized COMP models that retain the flexibility of these models while resolving the above issues. We use a Bayesian spatial filtering approach in order to efficiently handle high-dimensional spatial data and we use reversible-jump MCMC to automatically choose the basis vectors for spatial filtering. The COMP distribution poses two additional computational challenges-an intractable normalizing function in the likelihood and no closed-form expression for the mean. We propose a fast computational approach that addresses these challenges by, respectively, introducing an efficient auxiliary variable algorithm and pre-computing key approximations for fast likelihood evaluation. We illustrate the application of our methodology to simulated and real datasets, including Texas HPV-cancer data and US vaccine refusal data. Supplementary materials for this article are available online.

  PublisherOA PDFDOI

• Spatial distribution and determinants of childhood vaccination refusal in the United States

  Vaccine · 2023-04-15 · 4 citations

  articleOpen accessSenior author
  PublisherOA PDFDOI

• Fast Bayesian inference for spatial mean-parameterized Conway-Maxwell-Poisson models

  arXiv (Cornell University) · 2023-01-27

  preprintOpen accessSenior author

  Count data with complex features arise in many disciplines, including ecology, agriculture, criminology, medicine, and public health. Zero inflation, spatial dependence, and non-equidispersion are common features in count data. There are two classes of models that allow for these features -- he mode-parameterized Conway--Maxwell--Poisson (COMP) distribution and the generalized Poisson model. However both require the use of either constraints on the parameter space or a parameterization that leads to challenges in interpretability. We propose a spatial mean-parameterized COMP model that retains the flexibility of these models while resolving the above issues. We use a Bayesian spatial filtering approach in order to efficiently handle high-dimensional spatial data and we use reversible-jump MCMC to automatically choose the basis vectors for spatial filtering. The COMP distribution poses two additional computational challenges -- an intractable normalizing function in the likelihood and no closed-form expression for the mean. We propose a fast computational approach that addresses these challenges by, respectively, introducing an efficient auxiliary variable algorithm and pre-computing key approximations for fast likelihood evaluation. We illustrate the application of our methodology to simulated and real datasets, including Texas HPV-cancer data and US vaccine refusal data.

  PublisherOA PDFDOI

• Bayesian Spatial Models for Projecting Corn Yields

  Remote Sensing · 2023-12-23 · 2 citations

  articleOpen accessSenior authorCorresponding

  Climate change is predicted to impact corn yields. Previous studies analyzing these impacts differ in data and modeling approaches and, consequently, corn yield projections. We analyze the impacts of climate change on corn yields using two statistical models with different approaches for dealing with county-level effects. The first model, which is novel to modeling corn yields, uses a computationally efficient spatial basis function approach. We use a Bayesian framework to incorporate both parametric and climate model structural uncertainty. We find that the statistical models have similar predictive abilities, but the spatial basis function model is faster and hence potentially a useful tool for crop yield projections. We also explore how different gridded temperature datasets affect the statistical model fit and performance. Compared to the dataset with only weather station data, we find that the dataset composed of satellite and weather station data results in a model with a magnified relationship between temperature and corn yields. For all statistical models, we observe a relationship between temperature and corn yields that is broadly similar to previous studies. We use downscaled and bias-corrected CMIP5 climate model projections to obtain detrended corn yield projections for 2020–2049 and 2069–2098. In both periods, we project a decrease in the mean corn yield production, reinforcing the findings of other studies. However, the magnitude of the decrease and the associated uncertainties we obtain differ from previous studies.

  PublisherOA PDFDOI

• A Class of Models for Large Zero-inflated Spatial Data

  arXiv (Cornell University) · 2023-04-05

  preprintOpen accessSenior author

  Spatially correlated data with an excess of zeros, usually referred to as zero-inflated spatial data, arise in many disciplines. Examples include count data, for instance, abundance (or lack thereof) of animal species and disease counts, as well as semi-continuous data like observed precipitation. Spatial two-part models are a flexible class of models for such data. Fitting two-part models can be computationally expensive for large data due to high-dimensional dependent latent variables, costly matrix operations, and slow mixing Markov chains. We describe a flexible, computationally efficient approach for modeling large zero-inflated spatial data using the projection-based intrinsic conditional autoregression (PICAR) framework. We study our approach, which we call PICAR-Z, through extensive simulation studies and two environmental data sets. Our results suggest that PICAR-Z provides accurate predictions while remaining computationally efficient. An important goal of our work is to allow researchers who are not experts in computation to easily build computationally efficient extensions to zero-inflated spatial models; this also allows for a more thorough exploration of modeling choices in two-part models than was previously possible. We show that PICAR-Z is easy to implement and extend in popular probabilistic programming languages such as nimble and stan.

  PublisherOA PDFDOI

• A Shared Component Point Process Model for Urban Policing

  arXiv (Cornell University) · 2023-03-01

  preprintOpen accessSenior author

  Newly available point-level datasets allow us to relate police use of force to other events describing police behavior. Current methods for relating two point processes typically rely on the spatial aggregation of one of the two point processes. We investigate new methods that build upon shared component models and case-control methods to retain the point-level nature of both point processes while characterizing the relationship between them. We find that the shared component approach is particularly useful in flexibly relating two point processes, and we illustrate this flexibility in simulated examples and an application to Chicago policing data.

  PublisherOA PDFDOI

• Neglecting Model Parametric Uncertainty Can Drastically Underestimate Flood Risks

  Earth s Future · 2022-12-05 · 7 citations

  articleOpen access

  Abstract Floods drive dynamic and deeply uncertain risks for people and infrastructures. Uncertainty characterization is a crucial step in improving the predictive understanding of multi‐sector dynamics and the design of risk‐management strategies. Current approaches to estimate flood hazards often sample only a relatively small subset of the known unknowns, for example, the uncertainties surrounding the model parameters. This approach neglects the impacts of key uncertainties on hazards and system dynamics. Here we mainstream a recently developed method for Bayesian inference to calibrate a computationally expensive distributed hydrologic model. We compare three different calibration approaches: (a) stepwise line search, (b) precalibration or screening, and (c) the Fast Model Calibrations (FaMoS) approach. FaMoS deploys a particle‐based approach that takes advantage of the massive parallelization afforded by modern high‐performance computing systems. We quantify how neglecting parametric uncertainty and data discrepancy can drastically underestimate extreme flood events and risks. Precalibration improves prediction skill score over a stepwise line search. The Bayesian calibration improves the uncertainty characterization of model parameters and flood risk projections.

  PublisherOA PDFDOI

Recent grants

• Statistical Methods for Ice Sheet Projections using Large Non-Gaussian Space-Time Data Sets and Complex Computer Models

  NSF · $501k · 2014–2018

Frequent coauthors

• Klaus Keller

  Dartmouth College

  37 shared

• Won Chang

  Seoul National University

  26 shared

• Ben Seiyon Lee

  George Mason University

  25 shared

• Galin L. Jones

  20 shared

• David Pollard

  Pennsylvania State University

  17 shared

• Sanjib Sharma

  Howard University

  16 shared

• Iman Hosseini‐Shakib

  Pennsylvania State University

  15 shared

• Matthew J. Ferrari

  Pennsylvania State University

  14 shared