Significant contributor to Using Machine Learning Methods to Predict Treatment Outcome for Anxious Youth, National Institute of Mental Health (NIH), F31MH123038, awarded to Lesley Anne Norris, May 2020 - now
Goal: Use machine learning to analyze small sample data with many features and missingness. Mentoring Lesley Norris on machine learning, data collection, data cleaning and visualization.
Funded by EAGER: Assessing Influence of News Articles on Emerging Events, National Science Foundation, NSF-IIS-1842183, awarded to Obradovic, Z., Dragut, E, September 2018 - now
Goal: Analyze news articles and comments on emerging events, public opinions and persons.
Papers published: Stanojevic, M. et al., (2019, August); Stanojevic, M. et al. (2019, July), Alshehri, J et al. (in press)
Book chapter: Stanojevic, M., Alshehri, J., Obradovic, Z. (in press).
Other reports: Pham, Q., Stanojevic, M. and Obradovic, Z. (2020, May).
Awarded hardware through additional grant: Wrote an application for additional TACC supercomputer hardware resources with prof. Zoran Obradovic as a PI. I am managing the awarded resources since July 2020.
Phylogenetics Scientific Papers Understanding and Classification, January 2019 - now, with Institute for Genomics and Evolutionary Medicine (iGEM) Goal: Using NLP and information retrieval methods to select phylogenomics research papers containing timetree of rarely described species.
Funded by Pilot of a Cognitive Computing System to Analyze Immunization Data, September 2017 - May 2018, funded by Centers for Disease Control, subcontract to Abt Associates Inc.
Goal: Modeling complex data with textual and spatio-temporal dimensions. Using NLP techniques to understand documents and social media textual data. Technologies used: word2vec/Glove, NLP, Tensorflow, Python, Django, PySpark, Java, SQLite.
E-commerce Image Classification using Deep Learning, November - December 2017
Goal: Classifying user-provided images (~10 million) from e-commerce website into 5000 categories (ResNet, DenseNet, Python, TensorFlow, TfLearn).
Grocery Sales Forecasting, November - December 2017
Goal: Data integration and feature selection followed by PCA and linear, SVM and ANN regression (TfLearn and Scikit-Learn) to obtain best prediction.
Methylation Variation in Colon Cancer, April - May 2017
Goal: Understanding methylation variation in cancer patients using statistical modeling, complex networks and multi-layer clustering (Python, R). Working with little samples and a lot of features dataset.
Language detection using Deep Learning, July 2015
Goal: Learning RNN and LSTM models to detect language of wikipedia corpus (Python, Lasagne, Theano, Java).