<aside> 📌 PATH: 360° Customer View → Predictive Models
</aside>
How to ReTrain ML Model:
Consider retraining your ML model under the following conditions:
Accuracy Issues:
The current training results do not meet your business needs, or you aim to improve their accuracy and quality.
Decreasing Performance:
Predictions have become less reliable compared to when the model was initially deployed.
Availability of New Data:
You have access to new or updated data that can enhance the model’s training and prediction capabilities.
Changes in Market Dynamics:
Significant changes in market conditions, products, customer segments, or customer behavior necessitate incorporating these updates into the model to maintain relevance and accuracy.
Create datasets containing metrics and targets, and import them into Bind Genius.
If you wish to achieve different training outcomes, consider using alternative metrics and/or targets to test various model configurations.
The following articles may be helpful during this step:
How to Prepare Training Dataset with Metrics
How to Prepare Training Dataset with Targets
To evaluate whether your new data, metrics, or targets improve predictions, add and train a new ML model. This allows you to compare its performance with the current model effectively.