Exploratory Data Analysis (EDA)

EDA uses summary stats and visualisations to understand a dataset BEFORE modelling.

Ask: what's the distribution? Are there outliers? How do features relate to the target?

Example 1 (python)

import pandas as pd
df = pd.read_csv('data.csv')
print(df.describe())
print(df.info())

One-line dataset summaries.

Example 2 (python)

import seaborn as sns
sns.pairplot(df, hue='target')

Visualise feature pairs by class.

Key points

💡 Note: EDA is where you build intuition. Never skip it to jump straight to modelling — you'll pay later.