DAVOOD POURSINA

*** *. ******** ***. **********, OK 74075

405-***-**** https://www.linkedin.com/in/davood-poursina-b8547646/ aduhsm@r.postjobfree.com SUMMARY

10+ years of experience in analyzing the data using statistical methods

10+ years of experience in AB testing and design constructions

10+ years of experience in research and problem solving

5 years of experience in R programming

4 years of experience in ML

2 years of experience in SQL

TECHNICAL SKILLS

Programming Languages R, Stan, Python

Database SQL, pgAdmin, Rsql

Statistical and mathematical packages R, SAS, JMP, RSTAN, SPSS, GAMS, MAPLE, MINITAB Geo-spatial software Arc MAP, Geoda, MGWR

Statistical and mathematical methods T-test, Anova, Regression, Multivariate analysis, Design of Experiments (A/B testing), Casual Inference, Bayesian Modeling, Bayesian kriging, Time-series,

Mixes e ect models, Spatial Statistics, GLM

random coe cient model, SVC models

Optimization, Linear and Non-Linear programming

Dynamic programming

Machine Learning Algorithm (Supervised) Regression analysis, Logistic Regression,

(unsupervised) KNN, PCA, K-Mean, Random Forest,

WORK EXPERIENCE

Oklahoma State University, OK Aug. 2020 { Dec 2022 Graduate research Associate

Run a faster and feasible model for Bayesian spatially varying coe cient models using the sparsity of precision matrix instead of co-variance (2020).

Predict the missing values in data sets using the posterior prediction based on the spatial behavior of data instead of using the replacement methods (2020).

Find the WTP for each customer in a speci c area using the spatial behavior and probit models (2020).

Find the best treatment allocation for whole-farm experimentation in Linear to maximize the gained information from an experiment (2021).

Obtain the best treatment allocation for linear-plateau modeling (non-linear models) for on farm ex- perimentation to maximize the gained information (2021).

Combine the noisy and accurate measurements to guide the precision applications for agronomist (2022).

Conducting a fully Bayesian experimental design for On-farm experimentation in several consecutive years to maximize the farmers cumulative pro t (in progress). University of Kashan, Iran Sep. 2015 { Dec 2019

Assistant professor and Department head, Department of Statistics

Taught several applied statistical courses such as but not limited to Probability and statistics for engineers, Regression analysis I & II, Design of experiments I & II, Multivariate analysis, Categorical data analysis, Statistics for behavioral science.

Conducted multidisciplinary research with several departments.

Reduced the experimental runs for a nding the signi cant e ects and modeling the micro structure and mechanical property of steel-copper joint. I design an RSM design for nding the optimal values of factors based on a linear-quadratic model (2016-2017).

Conducted an especial fractional factorial split plot design for investigate the e ect of several factors, mixture of hard to change and usual factors, in a self-healing Bacterial air-entrained concrete and using the mixed e ect modeling to t a exible model (2017-2018).

Leadership, monitor managing the sta and scheduling the department program and syllabus. University of Kashan, Iran Sep. 2014 { Sep 2015

Assistant professor, Department of Statistics

Taught several applied statistical courses such as but not limited to Probability I&II, Design of exper- iments I & II, Sampling methods I&II.

Conducted multidisciplinary research with several departments. University of Isfahan, Iran Sep. 2007 { Sep 2014

Research assistant, Department of Statistics

Taught several applied statistical courses such as but not limited to Probability I&II, Design of exper- iments I & II, Statistics for economy and behavioral science.

Conducted and nding the best experimentation’s levels for small samples in generalized linear models

(GLM) with modi ed information matrix for small samples.

Found the pseudo-Bayesian optimal designs (robust experiments) based on a modi ed information.

Used the regression models for predicting the defects in roll-forming. EDUCATION

Oklahoma State University (OSU) Aug. 2020 -Expected Dec. 2022 Ph.D. in Agricultural Economics GPA: 4/4

Department of Agricultural Economics

Dissertation Topic: \Whole farm Experimentation: Making It Pro table" Supervisor: B. Wade Brorsen

Bayesian estimation in spatially varying coe cient models (SVC) is time-consuming and not com- mercially feasible. Therefore, this study employs the precision matrix’s sparsity to make these models signi cantly faster. In addition, it develops optimal experimental designs to maximize the gained information or the pro t on whole farm experimentation for an SVC model. University of Isfahan, Iran Sep. 2007 - Sep. 2014

Ph.D. in Applied Statistics

Department of Statistics

Dissertation Topic: \Design of experiment for generalized linear models" Supervisor: Hooshang Talebi

Classic experimental design criteria which approximate the estimators’ variance by the inverse of the Fisher information matrix are ine cient for experimentation with small sample sizes. This study develops adjusted information using the Bhattacharyya matrix to nd a precise approximation for the variance. As a result, a new class of modi ed optimal designs is obtained for generalized linear models.

University of Isfahan, Iran Sep. 2005 - Sep. 2007

MS in Applied Statistics

Isfahan University of Technology, Iran Sep. 2001 - Sep. 2005 BS in Statistics

CERTIFICATE

Udemy The Complete SQL Bootcamp 2022: Go from Zero to Hero SELECTED PROJECTS

Develop a fully Bayesian simulated based whole farm optimal design for maximizing the farmers’ pro t. OSU - Spring 2022

Develop a new method to nd the optimal allocations for treatment levels for whole farm experimentation for SVC models. OSU - Spring 2021

Code a faster and commercially feasible Bayesian non-linear model for spatial temporal model Using R and Stan. OSU - Fall 2020

Develop a theoretical new method for nding the site-speci c optimal nitrogen value for linear plateau models when both terms are random. OSU - Fall 2020 RESEARCH EXPERIENCE

PhD thesis project: \Whole farm Experimentation: Making It Pro table", Oklahoma State University at Stillwater, OK. Since Aug. 2020.

PhD thesis project: \Design of Experiments for Generalized Linear Models", University of Isfahan, Iran. Sep. 2007- Sep. 2014.

M. Sc. thesis project: \Design of Experiments for Generalized Linear Models", University of Isfahan, Iran. Sep. 2005 - Sep. 2007.

SELECTED PUBLICATIONS

Cho, W., ShalekBriski, A., Brorsen, B. W. and Poursina, D. (2022). \Combining low-cost noisy mea- surements with expensive accurate measurements to guide precision application". Precision Agriculture,1- 14.

Poursina, D., Brorsen, B Wade \ Where to Put Treatments for On-farm Experimentation". 15th Internationall Conference on Precision Agriculture. Minneapolis, USA 2022. Parastegari, N., Mosto nejad, D., and Poursina, D. (2022). \An Investigation on Bacterial Air- entrained Concrete Permeability Using Statistical Modeling". Advances in Civil Engineering Materials, 1-16.

Poursina, D. Brorsen, B Wade \Site-Speci c Nitrogen Recommendation; Using Bayesian Kriging Method with Di erent Correlation Matrices". Annual meeting AAEA and WAEA. Austin, USA 2021. Jafari, M., Abbasi, M., Poursina, D., Gheysarian, A., and Bagheri, B. (2017). \Microstructures and mechanical properties of friction stir welded dissimilar steel-copper joints". Journal of Mechanical Science and Technology, 31(3), 1135-1142..

Poursina, D., and Talebi, H. (2014). \Modi ed D-optimal design for logistic model". Journal of Statistical Computation and Simulation, 84(2), 428-437.

Contact this candidate