Rajendra Thapa

**** **** ** *** *

Charlotte, NC 28210

Phone: 740-***-****

Email: ad4l9u@r.postjobfree.com

Web: www.linkedin.com/in/rajendra-thapa

Google Scholar: Rajendra Thapa

Technical skills

• Data Modeling: Data mining, Exploratory Data Analysis, Regression Analysis, Classification (Support Vector Machines, Perceptrons and Kernels, Nearest Neighbor Methods, Naive Bayes, PCA, Random Forests, Ensemble Methods, Boosting Algorithms), Model validation and hyperparameter tuning, Natural Language Processing, Deep Neural Network, Recurrent Neural network, Recommendation systems, Machine Learning with Amazon SageMaker.

• Data Analysis and Visualization: SQL, Microsoft Power Tools, Matplotlib, Tableau, RStudio.

• Programming Languages: Python (Numpy, Scipy, Matplotlib, Pandas, Scikit-learn, Tensorflow), C++, R, JAVA, Octave.

• Simulation Techniques: Stochastic Processes, Monte Carlo, Molecular dynamics, Finite Element Analysis, Simu- lated Annealing

• Math/Statistics skills: Random number theory, Linear Algebra, Integral Calculus and Differential Equations, Probability, Population Sampling, Hypothesis testing, Probability distribution (Gaussian, Exponential, Poisson)

• General skills: LATEX, Linux, Shell-Script, High Performance Computing, Version Control (Git, GitHub), AWS, Microsoft Azure Fundamentals

Education

Ph.D. in Physics, Ohio University, Athens, Ohio, USA Aug 2017 - Jul 2022 M.S. in Physics, Ohio University, Athens, Ohio, USA Sep 2017 - May 2018 M. Sc. in Physics, Tribhuvan University, Kathmandu, Nepal Oct 2013 - Dec 2015 B. Sc. in Physics, Tribhuvan University, Kathmandu, Nepal Jan 2009 - Mar 2013 Experiences

Postdoctoral Researcher, Lehigh University Aug 2022 - Present

• Exploratory data analysis of big data obtained from material simulation using Python programming packages

(NumPy, SciPy, Pandas) and creating scientific visualizations using Matplotlib and Seaborn.

• Training a logistic regression with cross-validation to predict the time evolution of the class labels (crystal-like atoms and glass-like atoms) to study crystal growth rates using Python machine-learning packages (SciKit-Learn, Tensorflow, Keras).

• Using stochastic modeling to study the phenomena of nucleation, a stochastic process, during crystallization of glasses using SciKit-Learn.

• Using unsupervised machine learning (k-means clustering) to group atoms in different clusters based on their features and extend it to study time-evolution of cluster sizes using SciKit-Learn, Tensorflow. Research Assistant, Ohio University Sep 2017 - Jul 2022 Structure property relationships using regression:

• Used Python programming (NumPy, SciPy, Pandas, Matplotlib) for data mining large data-sets obtained from computer simulation to understand structure-properties relationship using techniques including, but not limited to, regression, classification, clustering, and time-series analysis.

• Combined classification algorithm with time-series analysis of atomic environment to predict the transition of coal waste to disordered graphite.

• Created a Python script from scratch that performs Monte Carlo simulations for material engineering Particle classification and time series analysis:

• Used random forest algorithm from Python machine learning package SciKit-Learn to identify the most important features from density, height, and diameter to successfully simulate amorphous carbon nanotubes formation

• Formulated a logistic regression algorithm to classify particles in liquids and glasses using the information about their surrounding as features (distances, angles, etc.). Validation and testing were done with data from neutron and x-ray scattering experiment.

• Use time-series analysis on displacement of particles to identify different class of atomic movement during liquid-glass transition

Experiment guided modeling of materials with Monte Carlo simulation:

• Updated and improved an algorithm that combined Markov Chain Monte Carlo technique and simulated annealing with conjugate gradient minimization to create realistic models of materials by creating various Python programming scripts

• Implemented this model to predict characteristic structures of various class of materials. Performance evaluation and fine tuning of the models were made using experimental results. Other projects: 2021-2023

• Build machine learning models with various algorithms (logistic regression, Bayesian Method, SVM, and Neural network and Deep Neural network) to predict credit risk using probability of default for German Cus- tomer data from UCI repository. Implemented a pipeline to combine imputation, standardization, dimension reduction, and model fitting and optimize hyperparameters through a cross-validated grid search.Github Link

• Build a pipeline for Parkinson’s disease detection using XGBoost Classifier. The feature selection were made using permutation feature importance and the performance of the model were compared for various combination of features.

• Build a convolutional neural network to identify celebrity images.

• Build a recommendation system to recommend music based on user browsing history. Special courses and certifications

• Certifications: 1) Machine Learning, 2) Google Data Analytics Professional Certificate, 3) Introduction to Data Science in Python.

• Ohio University: 1) Special Topics in Computer Science: Intro to C++, 2) Computer Simulation Methods, 3) Molecular Simulations, 4) Statistical Mechanics.

• Tribhuvan University: 1) Computational Physics, 2) Statistical Mechanics. Mentoring and Leadership

• Mentored incoming graduate students on material simulation and data analysis tools (May 2021- Jan 2022)

• Mentored students on simulation techniques at the Advanced Cyberinfrastructure Training for Modeling Physical Systems (Summer School) at Rensselaer Polytechnic Institute (June 2022, July 2023) Publications

• C.Ugwumadu, R. Thapa, Y. A. Majali, J. Trembly, and D. Drabold, phys. stat. sol. (b) 260, 2200527, 2023

• R. Thapa, C.Ugwumadu, K. Nepal, J. Trembly, and D. Drabold, Phys. Rev. Lett. 128, 236402, 2022

• R. Thapa, K. Prasai, R. Bassiri, M. M. Fejer, and D. Drabold, Phys. Rev. B 105, 224207, 2022

• R. Thapa, B. Bhattarai, M.N. Kozicki, K.N. Subedi, and D. Drabold, Phys. Rev. Mat. 4, 064603, 2020 Honors and awards

• NQPI Outstanding Dissertation Awards, 2022 : For the most influential dissertation at Ohio University.

• Vishwa S. Shukla Memorial Scholarship, 2020 : For academic and research excellence at Ohio University.

• NQPI Student Fellowship, 2019 : For research excellence at Ohio University.

Contact this candidate