a data analysis project that investigates the relationship between riding behaviour, weather conditions, and electric vehicle efficiency.
ride statistics are extracted from screenshots using ocr, enriched with historical weather data, and analyzed through exploratory visualizations and a linear regression model.
- ocr-based ride data extraction
- automatic weather data integration
- data cleaning and preprocessing
- exploratory data analysis
- correlation analysis
- linear regression for efficiency prediction
- python
- opencv
- easyocr
- pandas
- numpy
- matplotlib
- scikit-learn
- analyzed 500+ real-world rides
- achieved an r² score of 0.451
- mean absolute error of 1.53 wh/km for efficiency prediction