# Improve {% embed url="" %} Improve - Troubleshooting {% endembed %} ## Summary * The first step is to **address under-fitting:** * Add model complexity → Reduce regularization → Error analysis → Choose a more complex architecture → Tune hyper-parameters → Add features. * The second step is to **address over-fitting:** * Add more training data → Add normalization → Add data augmentation → Increase regularization → Error analysis → Choose a more complex architecture → Tune hyper-parameters → Early stopping → Remove features → Reduce model size. * The third step is to **address the distribution shift** present in the data: * Analyze test-validation errors and collect more training data to compensate. * Analyze test-validation errors and synthesize more training data to compensate. * Apply domain adaptation techniques to training and test distributions. * The final step, if applicable, is to **rebalance your datasets:** * If the model performance on the test & validation set is significantly better than the performance on the test set, you over-fit to the validation set. * When it does happen, you can recollect the validation data by re-shuffling the test/validation split ratio. --- # Agent Instructions: Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://fall2019.fullstackdeeplearning.com/course-content/training-and-debugging/improve.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.