Projects done:• LendingClub Default Borrowers Project: Based on csv file gathered from Github, this project builds a model that can predict whether a borrower will pay back their loan. Technical environment: SQL-Server, Shell, Google Colab, Docker, Kubernetes, Google Cloud, GitHub, Docker Hub Models used: Artificial Neural Network Model• Bank Authentication Project: Build a model that can predict whether a bank note is authentic. Technical environment: SQL-Oracle, Shell, Cloud Shell, RStudio, Docker, Kubernetes, Docker Hub, GitHub, Google Cloud, Terraform and AWS Models used: Artificial Neural Network Model• IBM-HR-Employee-Attrition-Analysis: Based on several variables, we try in this project to predict whether attrition is present in IBM-HR employees. Technical environment: SQL-Oracle, Shell, Cloud Shell, Google Colab, Docker, Kubernetes, Docker Hub, GitHub, Google Cloud, Terraform and AWS Models used: BaggingClassifier - Random Forest – AdaBoostClassifier -AdaBoost+RandomForest – GradientBoostingClassifier - XGBoostLogisticRegression RandomForestClassifier+AdaBoostClassifier• Boston Housing Price Project: This project tries to predict Boston Housing price based on multiple features. Models used: RANSACRegressor - Ridge - Lasso - ElasticNet – RandomForestRegressor – AdaBoostRegressor

Projects done:• Novartis Stock Time Series Analysis Project: Based on ten years daily time series data we try to predict the future two-week Novartis stock. Technical environment: SQL-Server, Shell, Cloud Shell, RStudio, Docker, Kubernetes, Google Cloud, GitHub, Docker Hub Model used : ARIMA – Exponential Smoothing - Neural Net• Biotech Stock Analysis Project: Based on two years daily time series data we try to predict the future one-week Novartis stock Technical environment: SQL-Server, Shell, Cloud Shell, RStudio, Docker, Kubernetes, Google Cloud, GitHub, Docker Hub Model used : Multi Layer Neural Net - ARIMA• USA Inflation Time series Analysis Project: After importing ten-year monthly data with no headers nor row IDs and in a sorted manner, the focus is to develop the forecast model for the next 24 months. Technical environment: Oracle-SQL, Shell, Cloud Shell, RStudio, Docker, Kubernetes, Docker Hub, GitHub, Google Cloud, Terraform and AWS Model used : ARIMA – Exponential Smoothing - Neural Net • UCI Salary Classification Project: With the UCI adult dataset, our task consists to predict if people in the dataset belong to one of the two class of salary. Technical environment: SQL-Server, Shell, Cloud Shell, RStudio, Docker, Kubernetes, GitHub, Google Cloud, Terraform and AWS Model used : Logistic Regression

Technical environment: Python and Google Colab• Data Import - Web Scraping, APIs & Wrappers Sources : Yahoo Finance, FOREX (FXCM) and EOD HISTORICAL DATA• Portfolio Optimization: o Doing the Mean-Variance optimization, finding the Efficient frontier and the Efficient (Market) portfolio, calculating Beta and classifying stocks between aggressive and not aggressive o Estimating the Securities Market Line and estimating Alpha parameter for short term analysis to determine the mispriced (overpriced and underpriced) stocks. o Estimating the CAPM and Fama & French Three-Factor and Five-Factor Models• Doing the Monte Carlo Simulations and Value-at-Risk (VAR) analysis to measure tail risk of stock return: We calculate the VAR based on Parametric Method, Historical Method as well as Bootstrapping Method. Then we calculate the Conditional Value-at-Risk (CVAR) to estimate the expected tail loss given the losses equal or greater than the value at risk.• Testing different Algorithmic Trading: o Trend-Following Strategy o Machine Learning-Based Strategy: We had used two technics:  Random Forest Classifier  Classification-Based Strategy o Back-testing Investment Strategies: The goal is to create active strategies that outperform a passive Buy-and-Hold Strategy for the Dow Jones and S&P 500 Indexes.  Momentum and Contrarian Strategy: In each strategy we take a long or short position based on certain conditions, going from simple to complex momentum or contrarian strategies. At the end of each strategy, we annualize the risk and return, then we compare that to those of the index.  Index Tracking and Forward Testing

• Data extraction via SQL Query to satisfy departmental data base users • Contribute to SQL Query Optimization• Ensure high level performance of SQL Query

• T-SQL Queries development leading to load OLTP data base• Elaborate data processing (ETL) via Visual Studio tools, SQL Server Management Services and Transact-SQL langages• MDX Queries development to process data with BI SSAS device and compute key performance indicators• Use of Enterprise Architect device

After creating Member, Title, Title_Copy and Reservation tables and insuring integrity constraints, I had created PLSQL Package leading to insert, modify and update members and films lists as well as to check films availability

- OLTP dimensional model creation.- Data Mart implementation from OLTP data base.- Initial data mart loading ( ETL SQL or SSIS process ).- Hypercube creation and management ( SSAS ).- KPI Development- Reports creation ( SSRS )

SQL Queries development to know doctors and medical devices availability relative to clients demand and the medical examination results. These information and examination results are stored in the data base