• Analyzed a dataset from Capital Bike Share in Washington, DC to learn about the relationship between the demand and the environmental factors in the area.

• Managed to come up with the final predictive regression model that can explain around 93.2% of the variability in the bikeshare demand.

Methods: Regression Analysis

Tools: R programming language

• Processed the raw image data.

• Visualized the differences of the classes of the satellite images such as trees, water bodies, etc.

• Recommended the next approaches to model the data.

Methods: Image Processing, Dimensionality Reduction, Data Visualization.

Tools: MATLAB

• Investigated the relationship between the fatty acid components of olive oil samples and their geographical location in Italy.

• Successfully identified 5 meaningful groups in the data based on the acid components. Produced a classification model with accuracy score of 91.86%.

Methods: Clustering Analysis, Dimensionality Reduction, Data Visualization.

Tools: Python - numpy, pandas, scikit-learn, matplotlib packages. 

• Analyzed Nike's web traffic data provided through Adobe Analytics Challenge 2020 competition using Adobe Analytics tool.

• Answered business questions and suggested solutions based on the insights drawn from the data.

Methods: Web Traffic Analysis, Data Visualization.

Tools: Adobe Analytics. 

• Used R to analyze NFL data compiled from the Pro Football Reference, which include all offensive football statistics at the team level for 10 seasons. The data has 350 observations with 26 continuous and 2 categorical variables.

• Produced a logistic regression model that has about 94.02% chance to distinguish if the team can make it to the playoff or not.

Methods: Regression Analysis, Data Visualization.

Tools: R programing language. 

• Used JMP to analyze a housing dataset in Ames, Iowa including 80 variables (23 nominal, 23 ordinal, 14 discrete, and 20 continuous) with 2930 observations.

• Managed to come up with the final predictive regression model that can fit around 90% of the data collected in the Ames Housing dataset.

Methods: Regression Analysis, Data Visualization.

Tools: JMP.