Pic
Machine Learning & Urban Computing
Pic

Mahdi Hashemi
Assistant Professor
Department of Information Sciences and Technology
George Mason University
Office: Engineering Building, Rm. 5353, Fairfax Campus
E-Mail: mhashem2@
PhD Institution: University of Pittsburgh
Google Scholar Profile

We live amidst real-time data flows, with sensors measuring everything from air quality to traffic, with our own cell phones and social media yielding information about our whereabouts and activity levels, with buildings reporting on their own energy consumption and maintenance. The role of machine learning in smart cities is to provide a seamless and data-driven operating environment. This is only one of the visions/goals of smart cities but one that is the focus of our research. Our research focuses on devising new algorithmic, statistical, mathematical, machine learning, artificial intelligence, and hybrid methods to tackle complex problems in urban computing and intelligent transportation, in addition to analytical and visualization methods to make sense of spatial-temporal data. We leverage the data to make cities smarter, environmental resources sustainable, and quality of life better for citizens.




Intelligent Transportation 2020 Spatial-Temporal Data Prediction 2020



How to Build a Normal Distribution Based on a Sample Normal Distribution and the 68-95-99.7 Rule
Z-Scores, Standardization, and the Standard Normal Distribution P Values
Hypothesis Testing Using P Values
Determinant Eigenvectors and Eigenvalues
Principal Component Analysis (PCA) Scale and Center the Data Before Principal Component Analysis (PCA)
Principal Component Analysis (PCA) in Python
Gradient Descent
Support Vector Machines (SVM) Support Vector Machines (SVM) in Python
Decision Trees Cost Complexity Pruning of Decision Trees
AdaBoost versus Gradient Boost AdaBoost
Decision Trees in Python XGBoost in Python
Introduction to Neural Networks Cross Entropy Cost Function versus Sum of Squared Residuals Cost Function for Neural Netwroks
Convolutional Neural Networks
Least Squares Regression Ridge vs Lasso Regression
Elastic-Net Regression (GLMNet) combines the Ridge Regression penalty with the Lasso Regression penalty
Bias and Variance Cross Validation
Confusion Matrix ROC and AUC
K-Means Clustering Hierarchical Clustering
K-Means Clustering versus Hierarchical Clustering
Smart Cities Smart Cities
Smart Cities Smart Cities Analytics
Video Analytics for Event Detection Intelligent Transportation Systems
Deep Reinforcement Learning for Traffic Light Control



One-Page Research Proposal Summary

At the beginning of every research, you need to write a one-page proposal and email it to me. This is a mandatory first step. It must include the following sections (font: Times New Roman, font size: 11, line spacing: Single, Paragraph spacing: 0):

Title
Title of your research.
Overview
Summary of what you want to do, why is it significant, what is the motivation for doing this research?
Contribution
What has been done in this field and how your work contributes in this field. Contribution means how your research would add to the state-of-the-art science or knowledge. Contribution does not mean describing what you like or want to do.
Impact
What pressing problem might it help us solve? What would the knowledge obtained from your research do for us? How does it advance science?
Challenges
What challenges you have to overcome to fulfill this research. Please note that by challnges I do not mean the steps that you have to take to address the research but difficulties that you have to overcome. For instance, coming up with a model that considers the temporal sequence of data because currently no such model exists.
Data
If your research requires data, from where and how you plan to obtain the data?

This is all assuming that you are not repeating someone else’s work. If someone else has already done something similar and answered these questions, you cannot duplicate it and claim it as your own.


Lab Unfinished Projects Excel Sheet

The following excel sheet helps me to keep track of who is working on how many and what projects: . This excel sheet is for you to put the title of your papers that are under development, under review, or rejected. You can see your name in this excel sheet. If you do not, just add your name. Write the title of each individual paper under your name. Before the title, determine the status of your paper. A paper’s status could be three things: (1) Under Development, (2) Under Review, and (3) Rejected n Times. For instance, if I am still working on a paper and it has never been submitted, I should write under my name:
Under Development: A new model for traffic prediction
Tentative Submission Date: Nov 30

If I have a paper that has been rejected two times and I am still working on it to fix it, I should write:
Rejected 2 Times: A new model for traffic prediction
Tentative Submission Date: Nov 30
If for any reason your tentative submission date changes, you should not wipe off the previous tentative submission date, but add the new date in front of the previous date, separated by a comma

If I have a paper that is under review in ICML2021, I should write:
Under Review in ICML2021: A new model for traffic prediction
Result Notification Date (for Conferences)/Submission Date (for Journals): Nov 30

If for any reason, the title of your project changes, you need to update the title in this excel sheet as well. When a paper is accepted, you must remove it from this excel sheet, because then it will go in your personal excel sheet under the published papers.
Whenever you submit a paper, you need to email me its original file: word or latex, along with PDF. Never email me only a PDF file of a paper.

Funding Policy for PhD

Following are the guidelines regarding how the PhD funding decisions are made:

Summer Funding:
To qualify for summer GTA, you have to have published or accepted a minimum of 3 papers during the past 12 months up to April 01.
Academic Year Funding:
To keep your academic year GTA funding for the next academic year, you have to have published or accepted a minimum of 1 paper during the past 12 months up to August 01. If the number of published papers is 0, you will not receive GTA funding for the next year.
To defend PhD proposal:
At least 5 papers (published or accepted), at least one of which should be a journal paper.
To defend PhD dissertation:
At least 10 papers (published or accepted), at least two of which should be journal papers.
Acceptable Publications:
Conferences with an acceptance rate of 24% or less. You can check their acceptance rate for previous years to get an estimate.
Journals must have an impact factor, be related to our field, and belong to a reputable publisher, such as ACM, IEEE, Elsevier, Springer, Wiley, and Talor&Francis.
Publication fees:
You are responsible for registration and travel costs associated with conferences and publication fees for journals. However, as a student, you qualify to apply for student travel awards from conferences and from our university.
The university might cover publication fees for some journals.

GMU Writing Center

The GMU writing Center (https://writingcenter.gmu.edu/) teaches writing skills for PhD students, e.g. dissertation writing. They offer one-to-one meetings. You are required to contact them and go through their training. I will check with you at some point to see your certificate of training through Mason Writing Center.


Graduate Student Research and Travel Grants

The office of Graduate Fellowships (gradfellows.gmu.edu) helps Graduate students find and win fellowships, scholarships, and grants. Contact Dr. Kay Agoston kagoston@gmu.edu to make an appointment.

List of Internal Awards: https://provost.gmu.edu/academics-and-research/graduate-education/awards-and-grants

List of external grants:


The Structure of Proposal for PhD Dissertation
  1. Abstract
    • In the Abstract, you should answer the question: how your research contributes "to promote the progress of science; to advance the national health, prosperity, and welfare; to secure the national defense; and for other purposes." The abstract should clearly highlight the contribution of your work.
  2. Chapter 1: Problem statement and motivation
    • Clearly state what problem is the focus of the work. State the research questions and/or hypotheses.
    • Why is it important to address the problem?
  3. Chapter 2: Introduction
    • Highlight the reasons why the problem needs to be addressed
    • Briefly describe the proposed approach
    • Explain the objectives
    • Explain the challenges
    • Explain the contributions
  4. Chapter 3: Background and related work
  5. Chapter 4: Proposed approach
  6. Chapter 5: Preliminary research
  7. Chapter 6: Proposed tasks/timelines


One PhD position is available, for applicants interested in the research conducted in our lab. Please submit your application for the IST PhD program here. The minimum requirements are:

Priority would be given to students who have conducted research under my supervision and have published with me.



He received the Ph.D. degree in Computing and Information from the University of Pittsburgh, PA, USA. He is currently an Assistant Professor with the Department of Information Sciences and Technology, George Mason University, Fairfax, VA, USA. His research focuses on Machine Learning and Urban Computing, where he also specializes in intelligent transportation, spatial-temporal data mining and prediction, and online social media data mining. He has published 34 journal articles and 15 conference papers. He served on the program committee of WI-IAT international conference and SEKE international conference and is a Reviewer Board Member of Information and a Topics Board Member of Remote Sensing, since 2020. He was nominated for the 2021, 2022, and 2023 George Mason University Teaching Excellence Award. He has developed and teaches machine learning, deep learning, information visualization, data analytics, and programming related courses, at both undergraduate and graduate levels.



Journal Papers

  1. M. Hashemi, "Geographical visualization of Tweets, misinformation, and extremism during the USA 2020 presidential election using LSTM, NLP, and GIS", Journal of Big Data, vol. 10, no. 125, 2023.

  2. M. Inoue, M.-H. Li, M. Hashemi, Y. Yu, J. Jonnalagadda, R. Kulkarni, M. Kestenbaum, D. Mohess, N. Koizumi, "Opinion and Sentiment Analysis of Palliative Care in the Era of COVID-19", Healthcare, vol. 11, no. 6, pp. 855, 2023.

  3. M. Inoue, M. Hashemi, M.-H. Li, R. Kulkarni, N. Koizumi, "Understanding the palliative care information circulating on Twitter during the Coronavirus pandemic", Innovation in Aging, vol. 6, no. Supplement_1, pp. 190-190, 2022.

  4. Z. Ara, M. Hashemi, "Predicting ride hailing service demand using Autoencoder and Convolutional Neural Network", International Journal of Software Engineering and Knowledge Engineering, vol. 32, no. 1, pp. 109-129, 2022.

  5. J. Jonnalagadda, M. Hashemi, "Feature selection and spatial-temporal forecast of Oceanic Nino Index using deep learning", International Journal of Software Engineering and Knowledge Engineering, vol. 32, no. 1, pp. 91-107, 2022.

  6. M. Hashemi, "Automatic type detection of 311 service requests based on customer provided descriptions", Applied Artificial Intelligence, vol. 36, no. 1, pp. 2073717, 2022.

  7. M. Hashemi, "Discovering social media topics and patterns in the Coronavirus and election era", Journal of Information, Communication and Ethics in Society, vol. 20, no. 1, pp. 1-17, 2022.

  8. M. Hashemi, "Studying and clustering cities based on their non-emergency service requests", Information, vol. 12, no. 8, pp. 332, 2021.

  9. M. Hashemi, "A data-driven framework for coding the intent and extent of political tweeting, disinformation, and extremism", Information, vol. 12, no. 4, pp. 148, 2021.

  10. M. Hashemi, "Forecasting El Niño and La Niña using spatially and temporally structured predictors and a convolutional neural network", IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 14, pp. 3438-3446, 2021.

  11. M. Hashemi, H.A. Karimi, "Weighted machine learning for spatial-temporal data", IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 13, pp. 3066-3082, 2020.

  12. M. Hashemi, M. Hall, "Multi-label classification and knowledge extraction from oncology-related content on online social networks", Artificial Intelligence Review, vol. 53, no. 8, pp. 5957-5994, 2020.

  13. M. Hashemi, "Web page classification: A survey of perspectives, gaps, and future directions", Multimedia Tools and Applications, vol. 79, no. 17, pp. 11921-11945, 2020.

  14. M. Hashemi, M. Hall, "Detecting and classifying online dark visual propaganda", Image and Vision Computing, vol. 89, no. 1, pp. 95-105, 2019.

  15. M. Hashemi, "Automatic inference of road and pedestrian networks from spatial-temporal trajectories", IEEE Transactions on Intelligent Transportation Systems, vol. 20, no. 12, pp. 4604-4620, 2019.

  16. M. Hashemi, "Enlarging smaller images before inputting into convolutional neural network: zero-padding vs. interpolation", Journal of Big Data, vol. 6, no. 98, 2019.

  17. M. Hashemi, H.A. Karimi, "Weighted machine learning", Statistics, Optimization and Information Computing, vol. 6, no. 4, pp. 497-525, 2018.

  18. M. Hashemi, M. Hall, "Visualization, feature selection, machine learning: Identifying the responsible group for extreme acts of violence", IEEE Access, vol. 6, no. 1, pp. 70164-70171, 2018.

  19. M. Hashemi, "Dynamic, stream-balancing, turn-minimizing, accessible wayfinding for emergency evacuation of people who use a wheelchair", Fire Technology, vol. 54, no. 5, pp. 1195-1217, 2018.

  20. M. Hashemi, "Emergency evacuation of people with disabilities: A survey of drills, simulations, and accessibility", Cogent Engineering, vol. 5, no. 1, pp. 1506304, 2018.

  21. M. Hashemi, "Reusability of the output of map-matching algorithms across space and time through machine learning", IEEE Transactions on Intelligent Transportation Systems, vol. 18, no. 11, pp. 3017-3026, 2017.

  22. M. Hashemi, "A testbed for evaluating network construction algorithms from GPS traces", Computers, Environment and Urban Systems, vol. 66, pp. 96-109, 2017.

  23. M. Hashemi, "Intelligent GPS trace management for human mobility pattern detection", Cogent Engineering, vol. 4, no. 1, pp. 1390813, 2017.

  24. M. Hashemi, H.A. Karimi, "Collaborative personalized multi-criteria wayfinding for wheelchair users in outdoors", Transactions in GIS, vol. 21, no. 4, pp. 782-795, 2017.

  25. M. Hashemi, H.A. Karimi, "A weight-based map-matching algorithm for vehicle navigation in complex urban networks", Journal of Intelligent Transportation Systems, vol. 20, no. 6, pp. 573-590, 2016.

  26. M. Hashemi, H.A. Karimi, "Indoor spatial model and accessibility index for emergency evacuation of people with disabilities", Journal of Computing in Civil Engineering, vol. 30, no. 4, pp. 04015056, 2016.

  27. M. Hashemi, A. Sadeghi-Niaraki, "A theoretical framework for ubiquitous computing", International Journal of Advanced Pervasive and Ubiquitous Computing, vol. 8, no. 2, pp. 1-15, 2016.

  28. M. Hashemi, H.A. Karimi, "A critical review of real-time map-matching algorithms: Current issues and future directions", Computers, Environment and Urban Systems, vol. 48, pp. 153-165, 2014.

  29. M. Hashemi, M.R. Malek, "Protecting location privacy in mobile geoservices using fuzzy inference systems", Computers, Environment and Urban Systems, vol. 36, no. 4, pp. 311-320, 2012.

  30. M. Hashemi, A.A. Alesheikh, M.R. Zolfaghari, "A spatio-temporal model for probabilistic seismic hazard zonation of Tehran", Computers & Geosciences, vol. 58, pp. 8-18, 2013.

  31. M. Hashemi, A.A. Alesheikh, M.R. Zolfaghari, "A GIS-based time-dependent seismic source modeling of Northern Iran", Earthquake Engineering and Engineering Vibration, vol. 16, no. 1, pp. 33-45, 2017.

  32. M. Hashemi, A.A. Alesheikh, "GIS: agent-based modeling and evaluation of an earthquake-stricken area with a case study in Tehran, Iran", Natural Hazards, vol. 69, no. 3, pp. 1895-1917, 2013.

  33. M. Hashemi, A.A. Alesheikh, "Development and implementation of a GIS-based tool for spatial modeling of seismic vulnerability of Tehran", Natural Hazards and Earth System Sciences, vol. 12, pp. 3659-3670, 2012.

  34. M. Hashemi, A.A. Alesheikh, "A GIS-based earthquake damage assessment and settlement methodology", Soil Dynamics and Earthquake Engineering, vol. 31, no. 11, pp. 1607-1617, 2011.

Conference Proceedings

  1. J. Jonnalagadda, M. Hashemi, "Quality-aware conditional generative adversarial networks for precipitation nowcasting", In 9th International Conference on Time Series and Forecasting (ITISE). Gran Canaria, Spain: MDPI, pp. 11, 2023.

  2. J. Jonnalagadda, M. Hashemi, "Optimizing the spatial-temporal extent of environmental factors in forecasting El Niño and La Niña using recurrent neural network", In 9th International Conference on Time Series and Forecasting (ITISE). Gran Canaria, Spain: MDPI, pp. 10, 2023.

  3. J. Jonnalagadda, M. Hashemi, "Long lead ENSO forecast using an adaptive graph convolutional recurrent neural network", In 9th International Conference on Time Series and Forecasting (ITISE). Gran Canaria, Spain: MDPI, pp. 5, 2023.

  4. Z. Ara, M. Hashemi, "Traffic flow prediction using long short-term memory network and optimized spatial temporal dependencies", In IEEE 9th International Conference on Big Data (BigData). Orlando, FL, USA: IEEE, pp. 1550-1557, 2021.

  5. J. Jonnalagadda, M. Hashemi, "A deep learning-based traffic event detection from social media", In IEEE 22nd International Conference on Information Reuse and Integration for Data Science (IRI). Las Vegas, NV, USA: IEEE, pp. 1-8, 2021.

  6. Z. Ara, M. Hashemi, "Identifying the severity of road accident impact on traffic flow by ensemble model", In IEEE 22nd International Conference on Information Reuse and Integration for Data Science (IRI). Las Vegas, NV, USA: IEEE, pp. 115-122, 2021.

  7. J. Jonnalagadda, M. Hashemi, "Spatial-temporal forecast of the probability distribution of Oceanic Niño Index for various lead times", In 33rd International Conference on Software Engineering & Knowledge Engineering (SEKE). Pittsburgh, PA, USA: KSI Research Inc., pp. 309-314, 2021.

  8. Z. Ara, M. Hashemi, "Ride hailing service demand forecast by integrating convolutional and recurrent neural networks", In 33rd International Conference on Software Engineering & Knowledge Engineering (SEKE). Pittsburgh, PA, USA: KSI Research Inc., pp. 441-446, 2021.

  9. J. Jonnalagadda, M. Hashemi, "Forecasting atmospheric visibility using auto regressive recurrent neural network", In IEEE 21st International Conference on Information Reuse and Integration for Data Science (IRI). Las Vegas, NV, USA: IEEE, pp. 209-215, 2020.

  10. P. Maktala, M. Hashemi, "Global land temperature forecasting using long short-term memory network", In IEEE 21st International Conference on Information Reuse and Integration for Data Science (IRI). Las Vegas, NV, USA: IEEE, pp. 216-223, 2020.

  11. S. Vadlamani, M. Hashemi, "Studying the impact of streetlights on street crime rate using geo-statistics", In IEEE 21st International Conference on Information Reuse and Integration for Data Science (IRI). Las Vegas, NV, USA: IEEE, pp. 231-236, 2020.

  12. M. Hashemi, M. Hall, "Identifying the responsible group for extreme acts of violence through pattern recognition", In International Conference on HCI in Business, Government, and Organizations. Cham: Springer, pp. 594-605, 2018.

  13. M. Hashemi, H.A. Karimi, "A machine learning approach to improve the accuracy of GPS-based map-matching algorithms", In IEEE 17th International Conference on Information Reuse and Integration (IRI). Pittsburgh, PA, USA: IEEE, pp. 77-86, 2016.

  14. M. Hashemi, H.A. Karimi, "Seismic source modeling by clustering earthquakes and predicting earthquake magnitudes", In Smart City 360°. Cham: Springer, pp. 468-478, 2016.

  15. M. Hashemi, A.A. Alesheikh, "Spatio-temporal analysis of Tehran's historical earthquakes trends", In Advancing Geoinformation Science for a Changing World. Berlin, Heidelberg: Springer, pp. 3-20, 2011.