Student

ARMIN KHAYYER

Data Scientist

[ *****.*******@*****.*** github.com/arminkhayyer www.linkedin.com/in/armin-khayyer/ Ó +1-703-***-**** R 425 Opelika Rd, Auburn, Al 36830 Bio: My Ph.D. research focuses on design and analysis of machine learning and statistical techniques in surrogate modeling and simulation optimization. My CSSE research is focused on state of the art Network embedding and Graph Alignment algorithms for large scale problems.

EDUCATION

Ph.D. Industrial and Systems Eng.

Auburn University

Aug. 2017 - present Auburn, AL

M.Sc. Computer Science and

Software Eng.

Auburn University

Aug. 2019 - present Auburn, AL

M.E. Industrial and Systems Eng.

Auburn University

Aug. 2017 - 2019 Auburn, AL

B.Sc. Industrial Engineering

Sharif University and Technology

Sep. 2012 - 2017 Tehran, Iran

COURSES

Data Mining, Adversarial Machine Learn-

ing, Dynamic Programming, Reinforcement

Learning, Adv. Algorithms, Operating Sys,

Linear Programming, Stochastic Program-

ming, Adv. Statistics, and Simulation Model-

ing and Analysis.

q PUBLICATIONS

• Kennedy, Joseph et al. (2020). “Efficient

Risk Estimation Using Extreme Value The-

ory and Simulation Metamodeling”. In:

Winter Simulation Conference.

• Khayyer, Armin et al. (2020). “Predict-

ing Public Transit Arrival Times: A Hybrid

Deep Neural Network Approach”. In: Jour-

nal of Big Data Analytics in Transportation,

Accepted.

B PROGRAMMING & SOFTWARE

Python Pytorch Tensorflow R

Keras scikit-learn OpenAI-GYM

NetworkX OpenCv Scipy NumPy

pandas Matplotlib Seaborn JS

Pyomo Django requests HTML

Css NGINX Gunicorn Docker

Google Cloud AWS Heroku

MySQL Shell Scripting LAT

EX VBA

C C++

EXPERIENCE AND RESEARCH

Research Assistantship

Auburn University

Sep. 2017 - Present Auburn, AL

•PredictiveModeling,GRA

Used state of the art Machine/Deep learning techniques, Kalman filtering, and statistical techniques to predict bus travel time and prevent bus bunching problems.

•Surrogate-Based Optimization and Metamodeling, GRA

- Developed an algorithm which improves the efficiency of kriging using clustering and parallel computing.

- Developed and implemented a simulation metamodel to estimate risk measures using extreme value theory.

•GraphMining,GRA

Used and modified state of the art algorithms such as Graph Convolutional Networks (GCN) for Network Embedding and further graph alignment purpose over huge datasets such as arXiv, dblp, and Acm.

Web Development

Auburn University

Sep 2018 - Present Auburn, AL

•Designedanddevelopeda research-based Web APP for human factor data collection and storage, analysis, and visualization. Wearable Dashboard

•Co-Designed and developed Driive Web APP which uses Google APIs, dynamic programming, and trajectory data mining techniques to optimize the fuel consumption.

ÿ SKILLS

•MachineLearning: Bias/variance, cross-validation, precision/recall, ROC curve, regularization, clustering (SOM, GMM), regression (Ridge, Lasso, linear, polynomial, logistic), PCA, PLS. Kriging, GPR, SVM

(Linear,poly,rbf), decision trees, ensemble learning (Random Forests), Bagging, and Stacking.

•Deep/Reinforcement Learning: MLP, Backprop, CNN, RNN, LSTM, GRU, Autoencoders (VAE), GAN, GCN, Graphsage, representation learning, Network Embedding, Seq2seq, Word2vec, attention, TD, Sarsa, Q-learning (SarsaMax), Deep Q-learning, and Policy Gradient.

•Optimization&Modeling: Convex, Integer, Nonlinear, Stochastic, and Dynamic programming, Constraint satisfaction, Multi-objective optimization.

- Iterative methods: Newton’s method, Sequential quadratic programming, Gradient descent, SGD, ADAM,and Quasi-Newton methods.

- Heuristics: Evolutionary algorithms, Genetic algorithms, Tabu search, Simulated annealing, Particle swarm optimization.

•WebandDatabaseDevelopment 4+ years professional experience with Django, SQL, HTML, JS, Css.

Machine/deep Learning

Optimization

Web and Database Development

Programming

Contact this candidate