Task 1: Data Cleaning & Preprocessing

🧠 Objective

The goal of this task was to clean and preprocess the stroke prediction dataset to make it suitable for machine learning.

🧰 Tools Used

• Python
• Pandas
• NumPy
• Seaborn / Matplotlib
• VS Code

📁 Files Included

• task1_preprocessing.py: All code used for loading, cleaning, encoding, standardizing, and saving the dataset.
• cleaned_stroke_data.csv: Final cleaned dataset.
• images/: Folder containing screenshots or plots (boxplots, missing value heatmap, etc.)

🧹 Preprocessing Steps

1. Load Dataset: Imported CSV using pandas.
2. Handle Missing Values: Filled missing bmi using median.
3. Encode Categorical Features: Used pd.get_dummies() with drop_first=True.
4. Standardize Numerical Features: Used StandardScaler from sklearn.
5. Remove Outliers: Used boxplots and IQR method.
6. Save Cleaned Data: Exported final dataset to cleaned_stroke_data.csv.

💾 Export Cleaned Dataset

The cleaned dataset was exported as a CSV file using pandas.DataFrame.to_csv(). You can find the final version in cleaned_stroke_data.csv.