Samy El Salamouny

Proven Data Analyst with a track record of leveraging Python programming and statistical analysis to drive decision-making. Excelled in transforming complex datasets into actionable insights, Skilled in SQL databases and adept at communicating complex findings, significantly improving data utilization for strategic planning.

Google Advanced Data Analytics Certificate - Coursera Platform

Providing data driven suggestions for HR (using python):

Input Data: A dataset provided by HR department in Salifort Motors organization of 14,999 rows &10 columns.

Goal: Analyzing the data collected by the HR department and building a model that predicts whether or not an employee will leave the company.

Result: After conducting feature engineering, the decision tree model achieved AUC of 93.8%, precision of 87.0%, recall of 90.4%, f1-score of 88.7%, and accuracy of 96.2%, on the test set. The random forest modestly outperformed the decision tree model.

Automatidata project (using python):

Input Data: A dataset provided by  New York City Taxi & Limousine Commission (New York City TLC) of 22,698 rows & 21 Columns. 

Goal: Predicting whether or not a customer is a generous tipper.

Result: Choosing Random Forest model as F1 score was 72.35% and it had an overall accuracy of 68.65%. It correctly identified 78% of the actual responders in the test set, which is 48% better than a random guess. It may be worthwhile to test the model with a select group of taxi drivers to get feedback.

POS Project (using python):

Input Data: Transactions & POS reports provided by Banque Misr.

Goal: Review unsettled POS transactions to get notifications about potential risks.

Result: Daily report of POS unsettled transactions & POSs that don't proceed transactions.

Medical Application Project (using python):

Input Data: samples collected from 253 doctors consists of 7 features and 1 label (the output target).

Goal: making predictions to classify if this doctor will write any of these drugs in prescription to his patients or not.

Result: choosing Decision Tree Classifier the best method to use as it gives the best F-score 83.33%.

Titanic Project (using python):

Input Data: collected data from the famous sunken ship “Titanic”.

Goal: making analysis to predict on the features of the survivors.

Result: Predictions with accuracy of 80.47% about features of survivors without using ML methods.

Finding Donors Project (using python):

Input Data: samples from potential donors.

Goal: using ML methods to predict on the income of potential donors,

Result: choosing AdaBoost Classifier method the best method to use as it gives the best accuracy score 87.02% & the best F-score 74.39%.

Transactions Dashboard (using Power BI):

Input Data: PTV transactions reports.

Goal: making analysis for decision making purpose.

Result: Transactions Dashboard.