in this notebook , i explored bike-sharing demand prediction using various regression models, including:

--------------------------------------------------------------------------------------------------------------------

1. Simple Linear Regression using temperature as the only feature.
2. Multiple Linear Regression using several meteorological and functional features.
3. Neural Network Regression, both:
   • A shallow model (single dense layer),
   • A deep model (multiple hidden layers with ReLU activations).

--------------------------------------------------------------------------------------------------------------------

steps :

1. Data Preparation:

• Load and clean the Seoul Bike Sharing dataset.
• Convert categorical variables to numeric.
• Filter data to a specific hour (noon) for temporal consistency.
• Drop weakly correlated or redundant features.

2. Exploratory Data Analysis:

• Scatter plots to visually inspect correlations between features and bike count.

3. Models Development:

• Implemented and trained models
• Normalization
• Loss function: Mean Squared Error (MSE).
• Optimizer: Adam with learning rate adjustments.

4. Evaluation & Visualization & Comparison:

• Comparison between linear and deep models using test MSE.
• Visualization of predicted vs. actual bike rentals.

--------------------------------------------------------------------------------------------------------------------

What was my Learning Objectives ? :

• Trying to understand regression modeling workflows.
• Apply data normalization and feature selection.
• Evaluate trade-offs between simple models and more complex neural networks.