90% of Kaggle Notebooks Include EDA: Is It Really Useless?

In a recent analysis of Kaggle notebooks, it was found that 90% of them incorporate Exploratory Data Analysis (EDA). EDA typically involves creating various visualizations such as plots, histograms, density functions, and scatter plots. Despite its widespread use, some argue that EDA is unnecessary. Critics point out that many of these notebooks ultimately employ advanced machine learning models like CatBoost or LightGBM. They claim that the EDA does not contribute to final decision-making processes. However, the prevalence of EDA in data science projects on platforms like Kaggle suggests that it remains a standard practice among data analysts and scientists.

Source: www.reddit.com