Post Job Free

Resume

Sign in

Data Scientist

Location:
Pearland, TX
Posted:
April 07, 2020

Contact this candidate

Resume:

Professional Experiences

Data Science Intern, Smart Wires Inc., San Francisco, California, CA. May 2019- Aug 2019

• Developed new machine learning API using deep reinforcement actor-critic learning on google cloud to solve the nonlinear optimization programming of company’s clients. Machine learning and Data Scientist Researcher, University of Houston, Houston, TX. Jan 2017-Present

• Conduct Bayesian Inference analysis to find the uncertainty and risks in cyber-attack detection systems using Hamiltonian Monte Carlo Markov Chain and Variational Beyesian Inference deep networks.

• Design novel variational sequence to sequence text generation platform to create sequence suggestion for academic papers. I used NLTK, Probabilistic Tensorflow APIs and combination of LSTM, GRU and Variational Auto-encoders structures.

• Conduct new self-learning algorithm based on deep residual actor-critic reinforcement learning and Monte Carlo tree search in smart energy systems. The method is based on AlphaGo Zero rules.

• Developed cyber-security methods using machine learning algorithms including generative adversarial deep networks, deep CNN classifiers. Tensorflow, Pytorch, Keras, SciKit-Learn, Scipy, Statsmodels and Pyspark are major machine learning packages used on local machines or clusters and clouds.

• Designed the optimal classifier selection and Time-series prediction python packages for signal processing. Implemented on IBM Watson clusters using Spark big data platform.

• Implemented new deep learning object detection methods including YOLO v2&v3 to find fault locations in electricity grids and improve system resiliency.

• Developed a novel Mixed Integer Programming model to find vulnerable cyber-attacks points in electricity grids, using mathematical programming with equality constraints. The simulations are implemented in MATLAB and GAMS. Software Developer, Mississippi State University, High Performance Computing Center, MS. Jan 2016-Dec 2016

• Created Corrective Action Planning toolbox (generation re-dispatch and load shedding algorithm) in PSSE (power system simulation software) using Python for ENTERGY Corp. Energy System Designer and Project Manager, Niroo Research Institute, Tehran, Iran. Sep 2012-Dec 2015

• Designed energy market analysis and prediction API, based on auto regressive integrated moving average models using SQL and Python.

• Implemented fault diagnosis of electricity distribution feeders in Tabriz province to find failure causes in electrical transformers and arresters. The project was simulated and tested in EMTP and ETAP electricity simulation software. Software Instructor, Amirkabir University of Technology, Tehran, Iran. Aug 2011-Dec 2015

• Instructing simulation and programming software including Python, SQL, MATLAB,, R, C++ and etc. Technical Skills

Programming languages: Python, SQL, MATLAB, C++, Java, FORTRAN,

Optimization: GAMS, CPLEX, GUROBI

Statistical Tools: R Studio, MINITAB

Parallel Computing: OpenMP, MPI, Spark, Hadoop, CUDA

Simulation Software: ETAP, PSSE, Digsilent, EMTP, OPAL-RT, Sematic S7

Additional: IBM Watson, Google Cloud, Linux, Amazon AWS Saeed Ahmadian

Cell: 713-***-****

Email: adco8n@r.postjobfree.com

https://github.com/saeedahmadian

https://www.linkedin.com/in/saeed-ahmadian88/

Summary of Qualifications

Electrical and Computer Engineering Ph.D. candidate experienced in data science and optimization.

• Deep Reinforcement Learning

• Time Series Analysis

• Machine Learning

• Data-driven Decision Making

• Natural Language Processing

• Programming (Python, SQL, C++)

• Statistical Data Analysis

• Big Data Analysis

• Text generation

Education

• Ph.D. Candidate in Electrical and Computer Engineering, University of Houston, Houston, TX. Dissertation: Cyber-security data-driven modelling of energy markets. GPA: 3.94/4.00

• M.Sc. in Electrical Engineering, Amirkabir University of Technology, Tehran, Iran. GPA: 3.90/4.00 Apr 2020

(Expected)

Sep 2013

Technical Projects

• Designed recurrent neural network (RNN) to predict stock prices using Keras package in python

(Compared with ARIMA model) and spark SQL to read data, (white paper for proposal).

• Implemented Optimal text classification in IBM Watson using Pyspark machine learning feature extraction and classification Algorithms including Logistic regression, Naïve Bayes, SVM, Decision tree and random forest, (workshop of NLP in University of Houston).

• Implemented Object detection method comparison between R-CNN, Fast R-CNN and YOLO V2 on Pascal Voc dataset, (Research group project).

• Implemented Deep style transferring using Convolutional Neural Networks, using TensorFlow library in Python, (workshop for Deep Learning course).

• Presented Optimal grid search classification and prediction methods with hyper-parameter tuning among different machine learning methods including LR, LDA/QDA, KNN, SGD, Logistic regression, Gaussian Naïve Bayes, Adaboost, decision tree, random forests, SVM, MLP, Using Scikit-Learn,

(Workshop for Deep Learning course and Data Science course).

• Compared different deep networks’ performance including AlexNet, VGGNet, ResNet, GoogleNet over different datasets, Automatic Learning & Data Mining course. Aug 2019

May 2019

Dec 2018

Sep 2018

Jan 2018

Sep 2017

Publications

Book:

• S. Ahmadian, “The Most Complete Applied Reference of ETAP”, NEGARANDEH DANESH publication, 664 pages,

(In Farsi), Dec 2013.

Journal and Conference Papers:

• S. Ahmadian, H. Malki, Z. Han, “Multi-Stage Dynamic Cyber-attack Mitigation Plan to Increase Smart Grid's Resiliency Using Stochastic Deep Policy Gradient reinforcement Learning”, ready for IEEE trans on Smart grids.

• S. Ahmadian, H. Malki, Z. Han, “Bayesian ARIMA Time-series to Predict Smart Grids Real-time Price Uncertainty Using both Hamiltonian Monte Carlo Markov Chain and Variational Bayesian Inference Networks”, submitted to WCCI.

• S. Ahmadian, H. Malki, Z. Han, “Detection of False Data Injection Attacks on Smart Grids Using Generative Adversarial Networks and Auto-Encoders”, IEEE transactions on systems man and cybernetics (waiting for publish).

• S. Ahmadian, H. Malki, Z. Han, “Cyber Attacks on Smart Energy Grids Using Generative Adversarial Networks”, IEEE GlobalSIP 2018 - 6th IEEE Global Conference on Signal and Information Processing.

• S. Ahmadian, J. Jahanipour, H. Malki, Z. Han, “Modelling Cyber Attacks on Electricity Market Using Mathematical Programming with Equilibrium Constraints”, IEEE Access, Feb 2019.

• R. Farajifijani, S. Ahmadian, S. Ebrahimi, E. Ghotbi “Wind Farm Layout Optimization Problem Using Joint Probability Distribution of CVaR Analysis” 2019 IEEE PES Innovative Smart Grid Technologies Conference.

• S. Ahmadian, H. Malki, AR. Sadat, “Modeling Time of Use Pricing for Load Aggregators Using New Mathematical Programming with Equality Constraints”, 2018 5th International Conference on Control, Decision and Information Technologies (CoDIT).

• AR. Sadat, S. Ahmadian, N. Vosoughi, “A novel torque ripple reduction of switched reluctance motor based on DTC- SVM method.” 2018 IEEE Texas Power and Energy Conference (TPEC).

• S. Ahmadian, H. Malki, M. Barati, “A novel demand response management model to reduce smart grid costs”, Poster presentation, University of Houston, GRaSP 2017.

• S. Ahmadian, B. Vahidi, J. Jahanipour, SH. Hosseinian, H. Rastegar, “A new price restricted optimal bidding model using derated sensitivity factors by considering risk concept”, IET Generation, Transmission & Distribution, Feb 2016. Certificates

• Applied AI with Deep Learning, Coursera online course.

• Advanced Machine Learning and Signal Processing, Coursera online course.

• IBM Big Data, Spark 1&2.

• High Performance Computing, University of Houston Data science center. May 2019

Apr 2019

Jan 2019

Sep 2018

Relevant coursework

Deep Learning Integer Programming Machine Learning Introduction to High Performance Computing Automatic Learning & Data Mining Cluster Computing Big Data Analysis Artificial Intelligence Methods in Smart Grids Energy Economics and Markets



Contact this candidate