How naeem1144/greatertorontoareahousepriceprediction works
Main Components
a) Data Layer:
• Input Data Source: "GTA_HomePrice_History.csv"
• Data features (Geographic, Sales, Listing, Market, Temporal, Building Type)
b) Data Preprocessing Pipeline:
• KNN Imputation for missing values
• One-hot encoding for categorical variables
• Feature scaling/normalization
• Train-test split
c) Model Layer (parallel branches):
• Branch 1: Linear Regression
• Branch 2: Decision Tree Regressor
• Branch 3: Gradient Boosting Regressor (with cross-validation)
• Branch 4: Artificial Neural Network (PyTorch)
• BatchNorm layers
• Early stopping mechanism
d) Evaluation Layer:
• Metrics calculation (R2, MSE, MAE)
• Cross-validation results
• Visualization components
Specific Elements
• Feedback loops for cross-validation and early stopping
• Input/output specifications at each stage
• Technology labels (Python, PyTorch, Scikit-learn)
Component Relationships
• Parallel model training paths
• Evaluation feedback loops
• Model comparison workflow
Technical Details to Emphasize
• Data preprocessing pipeline steps
• Model architectures (especially ANN layers)
• Cross-validation workflow
• Evaluation metrics flow