Hassan Ezzeddine
@linkedin.com/hassan-ezzeddine
Email : hasan-Mobile : -
Education
• American University of Science and Technology
Master of Science in Computer and Communication Engineering
Lebanon
Oct. 2017 – June. 2019
• American University of Science and Technology
Bachelor of Science in Computer and Communication Engineering
Lebanon
Oct. 2012 – Feb. 2016
Professional Development
• Machine Learning A-ZTM : Hands-On Python and R In Data Science
March. 2017
Udemy MOOC platform
• Deep Learning A-ZTM : Hands-On Artificial Neural Networks
June. 2018
Udemy MOOC platform
Experience
• Matic Services
Data Scientist
Lebanon
January 2020 - Present
◦ Machine learning: Implementing a machine learning model for predicting customer churn using R
◦ Dashboards: Building interactive dashboards using shinyapp in R and uploaded to rstudio connect, for several
marketing,sales and operations departments for accessing data.
◦ Automated Reports: Working on parameterized reports, and automated reports sent via email to employees in
charge
• Picosensors
Labview engineer / Data Scientist
USA
August 2019 - January 2020
I took part in a remote project ,where I used my knowledge in instruments measurement and simulation
software Labview, and created insights from data collected from air sensors.
◦ Remote measurements : Connecting to a Remote oscilloscope and applying frequency sweeping to function
generator for analysing frequency behaviour of measured signal from air sensors
◦ Data analysis and machine learning : Implementing machine learning models , through a batched
process.Where pollution is predicted in air and classifying the environment surrounding to a similar tested mediums
• Proximie inc.
Software Engineer
Lebanon
Aug 2016 - November 2019
◦ Speech Recognition: Working on a server side speech recognition solution in python (py-sphinx library),using an
open source project “CMU-Sphinx” a speech recognition engine .Also I have tried different solutions as amazon
Alexa and Microsoft azure speech recognition service.
◦ EMR: Implementing a standalone EMR (electronic medical record) system and integrating with epic app-orchard
using PHP , with descriptive data analysis for extracting some insights from patients data extracted from EPIC
EMR and caboodle database. Creating interactive visualizations using Tableau Public.
◦ Service integration: Developing the back-end services for a medical Surgeries system that unifies expertise and
Medicine student in one platform where surgeries are done using augmented reality,technologies used are Php and
Javascript
◦ Client and Server side scripts: Production, modification, and maintenance of website and web application user
interfaces using HTML,CSS and Javascript
• Data Aurora
Data Analyst
Lebanon
July 2014 - Aug 2016
◦ Data Analysis: Interpret data, analyze results using statistical techniques using python with pandas and numpy
libraries
◦ Data Visualization: Creating interactive data visualizations using Tableau public and Javascript (D3.js)
◦ Machine learning: Implementing machine learning models for data classification using k-nearest neighbor and
classification tree , using python and scikit-learn library
• Mentis Nation-Internship
Data Analyst
Lebanon
Jan 2013 - May 2013
◦ Data Visualization: Creating interactive timelines visualizations using Tableau and D3.js
◦ Data science tools: Getting familiar with data science tools applied on hands on projects using python
Research and Projects
I have high proficiency level in using python for the machine learning and data science projects I worked on. In
addition to the common daily used programming languages as Java-script and PhP in my current work.
• Ensemble learning for Stocks Prediction:: This is my current thesis project am working on, where it consists of
combining different type of analysis in the stock market, ranging from technical analysis , fundamental analysis , financial
news analysis and stocks network analysis. In the technical part all types of mathematical parameters were calculated
using Ta-lib package and created a new table from them and applied fed the to an ANN classifier for recognizing whether
it is an uptrend or downtrend .After data preprocessing stage and the model I applied some hyper-parameter tuning
using Gridsearch. In the fundamental part I used some apis using python to fetch companies fundamental data The
financial news part required data scraping using selenium module in python . And applied sentiment analysis over the
financial news headline using nltk module and applied a naı̈ve bayes machine learning model
• Neural Networks in Stocks: Comparison and implementing of BPA (back propagation algorithm multi-layer
perceptron) and a LSTM-RNN (long-short term memory recurrent neural network) in stock market. We described the
theory behind back propagation algorithm and recurrent neural networks, to be able to construct a stable program that
could learn from historical stock data, the future of given stocks.Models were implemented in python and used the
numpy, pandas, scikit learn and keras libraries in addition to the plotly library for visualizing results.
• Shiny app using R for financial dashboard :: The R-shiny app I worked on was used to visualise stock market
data as prices and ratios with predictions values of implemented ML models . It provides the user with clear comparison
between real values and predicted ones, or the indicators insights at a certain price.
• ECG signals analysis for Arrhythmia detection: Applying Neural Networks on ECG Signals of some patients to
check whether their heartbeats are normal or not, to predict the existence of Arrhythmia cardiac condition.Python
packages used for simulation are pandas ,scipy ,numpy ,and biosppy for the ecg signal processing before using them in
the neural network.
• Home Credit Default Risk-kaggle: Predicting how capable loan applicants of repaying a loan using ensemble
machine learning methods . And applied an intensive feature engineering best practices, and classifying users if they are
trustworthy , by analyzing there financial records. Support vector machine (SVM) model was implemented for
classification in python with scipy module.
Technical Skills
• Data science and Machine Learning Algorithms: Data Preprocessing,
Regression,Classification,Clustering,Dimensionality Reduction( PCA, LDA, Kernel PCA),Model Selection Boosting(
k-fold Cross Validation, Parameter Tuning, Grid Search, XGBoost) in addition to Artificial ,Convolutional and Recurrent
Neural Networks. Backpropagation algorithm
• Coding skills: Proficient in Python,C++,Javascript,PHP and R, all the experience gained in those programming
languages was from my current work at proximie and personal effort when working on projects in university and
freelancing.
• Open source libraries: Pandas,Scikit Learn,Keras,matlplot,plotly
Prizes and Awards
• Beirut Startup Weekend (Artificial intelligence edition) Competition won the third Prize for an
artificial intelligent personal trainer idea.
• Numerical methods Competition won the first Prize for in the university for the numerical methods
competition where we implemetened a suduko solver using python.
