Book Review — Building Machine Learning Powered Applications

Book Review — Building Machine Learning Powered Applications by Emmanuel Ameisen

I found this book a bit frustrating to follow.

Wherever the code is shown in the book and a link provided — it is to the main github repo and not to the individual jupyter notebook or python code. Nor was the filename mentioned beneath the code snippet.

Not a deal breaker but it is frustrating.

Pros

• Covers a lot of theory on building a machine learning application
• You build a ML Editor that gives recommendations
• Important concepts are illustrated in specific notebooks
• Incremental changes to the model for better results

Cons

• More theory, less code
• Light on deployment details
• Book links point only to the main repo and not the specific file with the code
• Not all tied together (filter models / tests).

My github repo — It contains the chapter numbers in the filenames.

Some images from the final built machine learning app with recommendations.

1. Main page of ML Editor

2. Model V1 — Sample Question

3. Model V1 — Readability score

4. Model V2 — Sample Question

5. Model V2 — Question Score

6. Model V3 — Sample Question

7. Model V3 — Detailed Recommendations

8. Model V3 — Another Recommendation example (Copied from writers.stackexchange.com)

Originally published at https://dev.to on January 12, 2021.