Access this on-demand session to learn practical methods for improving the accuracy of predictive models using IBM SPSS Modeler. While the demonstrations use Modeler, the techniques discussed are widely applicable and relevant to anyone working with predictive analytics, regardless of the software they use.

The session explores six proven approaches to enhancing model performance: bootstrap aggregation, boosting, feature engineering, ensemble methods, meta-modelling and split-method modelling. Each technique is explained in clear, practical terms, with examples that show how these methods can produce more robust and reliable predictions.

Whether you are refining existing models or looking to extend your analytical toolkit, this session provides a focused overview of strategies that can significantly improve predictive accuracy.

• Bootstrap aggregation
• Boosted models
• Feature engineering
• Ensemble models
• Meta-modelling
• Split-method models