Selecting highly prospective locations for data collection projects by designing and deploying classification algorithms including random forest and SVM
The notebook provides a step-by-step guide to preparing and analyzing geospatial data and creating a potential map using supervised ml techniques:
- Data cleaning and re-formatting
- Data standardization
- Missing value imputation
- Exploratory data analysis
- Data visualization
- Principal component analysis and descriptive statistical analysis
- Feature generation
- Customized trainig and test data splitting
- Support vector classifier and random forest algorithm training
- Hyper-parameter tuning
- Evaluating model performance
- Deployment of trained model on a new dataset
- Generate visualization products