Group 2 Training, Evaluation and Optimization

Tasks for Group 2 in KNIME Data Science Learnathon
- Train a Decision Tree on the training set, and apply the model to the test set
- Evaluate the performance of the Decision Tree model
- Train a Logistic Regression model on the training set, and apply the model to the test set
- Evaluate the performance of the Logistic Regression model
- Optimize the tree depth of a Random Forest model, and train and apply a Random Forest model using the optimal parameter value
- Evaluate the performance of the Random Forest model
- Compare the performances of the different models using scoring metrics for a classification model and an ROC Curve
- Write the best performing model to a file

External resources

• KNIME Analytics: a Review
• Building a Basic Model for Churn Prediction with KNIME
• Model Selection and Management with KNIME
• Behind the Scenes of Decision Tree with KNIME
• Decision Tree Learner Node: Algorithm Settings
• Ensemble Learning
• Import Existing Models
• KNIME E-Learning Course - Predictive Analytics
• From Modeling to Scoring: Confusion Matrix and Class Statistics
• Scoring Metrics for Classification Models
• Cross Validation with SVM
• Cross-validation (statistics)
• Parameter Optimization for Prediction Models
• Analytics - Model Selection to Predict Flight Departure Delays
• Original Airline Dataset