Sahel Eskandar

Data Scientist with 6+ yrs. of blended industrial and academic experience. Expert in data analysis and statistical analysis. Problem solver experienced using open source libraries and applying a wide range of machine learning techniques. Seeking a challenging environment and problems to solve and grow within a team.

• Data Preprocessing (noise reduction, alignment, rebalancing, encoding) & Exploratory Data Analysis
• ML algorithms design and implementation (Ensemble Learning, kNN, Gradient Boosting, Decision Trees, and Naive Bayes)
• Effectively advanced a multiclass classification model; increased accuracy by 12% (0.98 F1 scores)
• Successfully developed unsupervised Anomaly Detection using TensorFlow to predict multivariant time series instant anomalies within 1.25 sec
• Established a supervised data-driven classification approach using adaptive neuro-fuzzy inference systems resulting in lower redundancy and 16 times faster processing
• Data collection design, implementation, and integration from multiple sensors
• Collaborated with stakeholders to identify the business problems
• Industrialized tools to extract information from Autodesk caused 30% less time to prepare a report
• Extracted insights and improved tracking progress via Python and SQL
• The synchronized system resulted in 5x faster data management and 50% fewer assigned resources

Doctorate in Engineering at McMaster University

Specialize in Deep Learning and Applications, Neural Networks and Development Tools, Combinatorial Optimization, Systems Engineering, Multivariate Statistical Models and Data Processing Methods