Ramin Ghorbani

Education

• 2020 - Present
  Ph.D. Researcher, Delft University of Technology, Delft, Netherlands
  Pattern Recognition & Bioinformatics
  Project: STRAP: Self TRAcking for Prevention and diagnosis of heart disease
  Thesis: Self-Supervised Representation Learning for better learning in label-scarce regimes
• 2017 - 2019
  Master of Science, Iran University of Science and Technology, Tehran, Iran
  Systems Optimization (Data Mining in HealthCare)
  Thesis: A new hybrid predictive model to predict the early mortality risk in Intensive Care Units on a highly imbalanced dataset
  GPA: 3.88/4
• 2013 - 2017
  Bachelor of Science, Yazd University, Yazd, Iran
  Industrial Engineering (Data Analysis)
  Major GPA: 3.12/4

Publications

• Comparing different resampling methods in predicting students' performance using machine learning techniques
  Ramin Ghorbani, Rouzbeh Ghousi
  Journal of IEEE Access - (Published 2020)
• A new hybrid predictive model to predict the early mortality risk in Intensive Care Units on a highly imbalanced dataset
  Ramin Ghorbani, Rouzbeh Ghousi, Ahmad Makui, Alireza Atashi
  Journal of IEEE Access - (Published 2020)
• Predictive data mining approaches in medical diagnosis: A review of some diseases' prediction
  Ramin Ghorbani, Rouzbeh Ghousi
  International Journal of Data and Network Science (Published 2019)
• Location of compressed natural gas stations using multi-objective flow refueling location model in the two-way highways: A case study in Iran
  Ramin Ghorbani, Rouzbeh Ghousi, Ahmad Makui
  Journal of Industrial and Systems Engineering (Published 2019)

Experiences

I have been extensively involved in different projects, which has allowed me to improve my personality, abilities, and technical skills, including implementing state-of-the-art Machine Learning and Deep Learning frameworks. I am energetic, open-minded, and actively strive for collaboration.

Research Assistance

Delft University of Technology (TU Delft), Delft, Netherlands

• Representation Learning / Learning from Few Labeled Samples (2020 - Present)
  Working on representation learning methods in detail while focusing on Self-Supervised Learning method in order to have an informative representation that can be used for different downstream tasks using few numbers of labeled samples


• Self-TRAcking for Prevention and diagnosis of heart disease (2020 - Present)
  Collaborating with colleagues from other universities on development and implementation of a Machine Learning framework in order to detect heart deterioration patterns in elderlies population using PPG clinical signal from Smartwatches


• Time-Series Analysis and Signal Processing (2020 - Present)
  Learning different preprocessing steps for this type of data and trying to improve reconstruction and forecasting performance using different Machine Learning and Deep Learning method

Iran University of Science and Technology, Tehran, Iran

• Handling Imbalanced Datasets (2019)
  Focusing on different methods of handling imbalanced datasets while trying to get acceptable and perfect performance using different Machine Learning methods


• Predictive Machine Learning models for Health Data (2019)
  Development and implementation of classic Machine Learning framework models on tabular health data such as ICU mortality dataset

Teaching Assistance

• Machine Learning 1 Course / Fall 2022
  My responsibilities include providing technical support for lectures and assigments

• Machine Learning Basics Course / Spring 2019
  My responsibilities include lectures, exams, and homework assignments

Selected Projects

• Developing a machine learning framework using sensor data collected from smartwatches to distinguish between HF patients (In Progress)
  This project aims to uncover the structure of clinal PPG signal in order to classify acute decompensated HF at hospital admission, admitted HF patients before discharge, stable HF from the outpatient clinic, and healthy volunteers detect colorectal cancer in the early stages and reduce the probability of developing colorectal cancer.


• Early detection of colorectal cancer using different machine learning models (Accomplished)
  This project aims to detect colorectal cancer in the early stages and reduce the probability of developing colorectal cancer.


• Prediction of best fertility treatment using machine learning techniques (Accomplished)
  This project tries to predict the best fertility treatment for infertile couples using various machine learning models.


• Using data mining techniques for Customer Relationship Management (Car Factory in Iran) (Accomplished)
  This project uses data mining techniques to identify valuable customers, predict future behaviors, and detect the cost of the services based on different features.

Contact

Email: R.Ghorbani@Tudelft.nl

Address:
TU Delft, building 28, room 6.E.200
Van Mourik Broekmanweg 6
2628 XE Delft
The Netherlands

• R^G
• GitHub

Elements

i = 0;

while (!deck.isInOrder()) {
    print 'Iteration ' + i;
    deck.shuffle();
    i++;
}

print 'It took ' + i + ' iterations to sort the deck.';