Example 2Lesson 30Free preview

Ex2 Training

This lesson is part of A.I. · All levels

Summary

Once the house price model is ready, the training is run with normalized data, often with a particular batchSize so that weight updates occur more frequently. After training, the new house can be entered into the model.predict() to obtain a price estimate. Evaluation of the prediction results determines whether the model needs to be improved.

Key points

BatchSize determines how many data samples before the weight is updated.
Training with epoch is enough to make the loss decrease to convergence.
New home predictions should be normalized in the same way as data training.
The predicted results compared to the actual price to assess the quality of the model.
The resulting inconsistencies become the material for data improvement and model architecture.

Full lesson

Code examples and quizzes

Full lessons, code examples, and quizzes are available to subscribers. 5 questions.