About Me
Know Me More
I'm Mojtaba Heidarysafa, a Machine Learning Engineer and Data scientist
Since 2016, I worked with Professor Donald E. Brown (Founding Director of the Data Science Institute at UVA) to use data science toolbox for various of applications. During my Ph.D at University of Virginia, I developed new machine learning and deep learning models, published papers in top-tier Conferences, and used data science solutions for real-world applications . I have a solid understanding of machine learning algorithms, deep learning, stat and probability concepts as well as programming.
My interests include Natural language Processing, Data-driven solutions, supervised/unsupervised machine learning, forcasting, and artificial intelligence. More information is available at my projects and publication sections below .
- Name:Mojtaba Heidarysafa
- Email:mh4pk@virginia.edu
Pojects
My Work
Publications
Selected Publications
Text classification algorithms: A survey Journal of Information (2019)
A comprehensive survey of methods for pre-processing textual data as well as algorithms used for text classification.
A propsed hierarchical structure using two levels where each level can be either of neural nets, recurrent nueral nets, or multilayer perceptron. The model was successfully implemented for text classification task of scientific papers.
Women in ISIS propaganda: A natural language processing analysis Science and Information Conference (2020)
This paper analyzed a corpus of ISIS propaganda materials while comparing it with main stream religious group. Best topics were found by non negative matrix factorization in both groups. Invoked emotion analysis was performed using Depechemood (a lexical-base emotion analysis method)
An improvement of data classification using random multimodel deep learning (rmdl) International Journal of Machine Learning and Computing
This work presents a bagging of classifiers using main deep learning architectures such as multi layer perceptron, convolutional neural networks, and recurrent neural nets. The proposed model shows great performance on image and text classification tasks.
This paper presents using word embeddings and deep learning architectures to successfully detec the cause of a railway accident based on the accident description.
Complete list of my current publications as indexed by google which includes all other publications not mentioned here.
Summary
Resume
My Education
2016 - present
Doctor of Philosophy
University of Virginia
Solving real world problem using machine learning/deep learning and natural language programming methods
2015
Master of Science
Tampere University of Technology
Building understanding of robotics challanges and programming both atonomous robots and industrial manipulators
My Experience
2021
Natural Language Processing Lecturer
School of Data Science (University of Virginia)
Offering a course on fundamental of NLP, text classification, and new advancements in deep learning for text as data
2019
Data Science Fellowship
School of Data Science (University of Virginia)
Bringing natural language processing techniques to other domains such as politics and social media analysis
Skills
Python Programming: numpy, pandas, scikit learn, keras , pytorch 90%
Programming Languages: R, Java, Matlab 70%
Theory: Machine Learning, statistics, NLP, time series 95%
Neural Nets: MLP, CNN, RNN, LSTMs, GRU, Transformers, BERT85%
Cloud : AWS (sagemaker, lambda, EC2, CloudWatch) 80%
database: SQL, MySQL 70%
Other Technologies: Dash, Bokeh, Tableau, spark 60%
HTML/CSS/Web Programming 50%
Contact
Get in Touch
Address
Thornthon Hall
Engineering Way
Charlottesville VA
(703) 656 6964
mh4pk at virginia.edu