add normalization step

🛠️ Description of Contribution: Added Feature Normalization Step

🔄 What I Did:

I added a feature normalization step to the dataset preprocessing pipeline in the implementation of the machine learning algorithms (Linear regression & Logistic regression ). The original repository did not include this step, which can significantly impact the performance and convergence behavior of many ML models.

📌 Summary of Changes:

• Introduced a normalization function that scales input features to a standardized range (mean = 0, std = 1) or [0, 1], depending on the model requirement.
• Updated the training pipeline to apply normalization before model fitting.
• Ensured that the same transformation is applied during inference/prediction.
• I did not extract normalization into a shared module. This ensures clarity and local control for each model’s behavior.

📈 Why Normalization Matters:

Feature normalization is a crucial step in machine learning, especially when:

• Features have different units or scales (e.g., age in years vs. income in dollars).
• Models rely on gradient-based optimization (e.g., Linear Regression, Logistic Regression, Neural Networks), where unscaled features can lead to slow convergence or divergent gradients.
• We want fair contributions from each feature without one dominating due to larger magnitude.

Without normalization:

• The cost function may not converge efficiently.
• The model might be biased toward high-magnitude features.
• Training can become unstable or yield suboptimal results.