Analyzed 3M+ records of 2018 bike-sharing data of Washington city, identified the different user riding patterns between registered and casual customers, analyzed the key factors of converting casual users to membership. Applied techniques include SAS macro and logistic models.
Technical Objective To predict rider’s membership type by analyzing ride sharing records. Business Objective To increase customer loyalty and market share therefore achieve more sales.
Data Collection Source data provided by Metro College of Technology. 2. Data Definition 2018 Washington City bike sharing records. 3. Data Scope 12 .csv files, 9 variables, 3.5 million observations. 4. Software Used SAS 9 5. Statistics Methods Used i. Data Cleaning Handle missing values ii. Feature Engineering Handle outliers iii. EDA Univariate & Bivariate Analysis, Hypothesis test iv. Data Visualization Histogram, Bar Chart, Pie Chart, Boxplot v. Logistic Regression