Reader small image

You're reading from  Developing Kaggle Notebooks

Product typeBook
Published inDec 2023
Reading LevelIntermediate
PublisherPackt
ISBN-139781805128519
Edition1st Edition
Languages
Right arrow
Author (1)
Gabriel Preda
Gabriel Preda
author image
Gabriel Preda

Dr. Gabriel Preda is a Principal Data Scientist for Endava, a major software services company. He has worked on projects in various industries, including financial services, banking, portfolio management, telecom, and healthcare, developing machine learning solutions for various business problems, including risk prediction, churn analysis, anomaly detection, task recommendations, and document information extraction. In addition, he is very active in competitive machine learning, currently holding the title of a three-time Kaggle Grandmaster and is well-known for his Kaggle Notebooks.
Read more about Gabriel Preda

Right arrow

Revisit and refine your work periodically

When I create a notebook, it is quite unusual to just put it aside and then start working on a new topic. Most of the time, I will return to it several times and add new ideas. In the first versions of the notebook, I try to focus on data exploration and really understand what is uniquely characteristic about the respective dataset (or datasets). In the next versions, I work on refining the graphics and extract maybe functions for data preparation, analysis, and visualization. I organize the code better, eliminating repetitive parts and eventually saving the generic parts in a utility script. The best part of using utility scripts is that you now have reusable code that can be used in multiple notebooks. When I create a utility script, I take steps to make the code more generic, customizable, and robust.

Next, I refine the visual identity of the notebook as well. I check the unity of the composition, making changes to the style to adapt...

lock icon
The rest of the page is locked
Previous PageNext Page
You have been reading a chapter from
Developing Kaggle Notebooks
Published in: Dec 2023Publisher: PacktISBN-13: 9781805128519

Author (1)

author image
Gabriel Preda

Dr. Gabriel Preda is a Principal Data Scientist for Endava, a major software services company. He has worked on projects in various industries, including financial services, banking, portfolio management, telecom, and healthcare, developing machine learning solutions for various business problems, including risk prediction, churn analysis, anomaly detection, task recommendations, and document information extraction. In addition, he is very active in competitive machine learning, currently holding the title of a three-time Kaggle Grandmaster and is well-known for his Kaggle Notebooks.
Read more about Gabriel Preda