Define the experiment
Choose a prediction problem and measurable success criteria.
Have AI prepare a guided notebook and explain its stages.
Build tensorflow experiments projects with AI assistance. Describe the need, guide the tools, test the results and improve the solution with a trainer. Technical concepts are explained through the project rather than memorised syntax.
Choose a prediction problem and measurable success criteria.
Have AI prepare a guided notebook and explain its stages.
Ask AI to organise a demonstration dataset.
Check data quality and the separation of training and test examples.
Have AI run guided experiments and compare them with a simple baseline.
Inspect incorrect predictions and check the evaluation metrics.
Request targeted changes and compare the measured results.
Check overfitting, new examples and conditions where the model fails.
Have AI prepare a demonstration interface.
Present the tests, results and limitations with the trainer.
Ready to build a verified project with AI?