House Price Prediction Model
Model Summary
A regression model that predicts residential house prices based on structured tabular features such as square footage, location, number of rooms, and other property attributes. Built as part of a broader machine learning portfolio exploring supervised learning across regression tasks.
Model Details
- Developed by: Chandrasekar Adhithya Pasumarthi (@Adhithpasu)
- Affiliation: Frisco ISD, TX | AI Club Leader | Class of 2027
- Model type: Regression (Neural Network / Gradient Boosting — update as applicable)
- Framework: TensorFlow / Keras (or scikit-learn — update as applicable)
- License: Apache 2.0
Intended Uses
Direct use:
- Predicting house sale prices from structured property data
- Exploring feature importance in real estate pricing
- Educational demonstrations of regression pipelines and feature engineering
Out-of-scope use:
- Production real estate valuation without domain expert review
- Generalization to housing markets with significantly different distributions than training data
Training Data
Trained on a structured housing dataset containing features such as:
- Square footage (total, above ground, basement)
- Number of bedrooms and bathrooms
- Neighborhood / location
- Year built and year remodeled
- Garage, pool, and lot features
(Update with your specific dataset — e.g., Kaggle's Ames Housing Dataset)
Evaluation
| Metric | Value |
|---|---|
| RMSE | TBD |
| MAE | TBD |
| R² | TBD |
(Fill in with your actual test set results)
How to Use
import tensorflow as tf
import numpy as np
# Load model
model = tf.keras.models.load_model("house_price_model")
# Example input — replace with your actual feature vector
# [sq_ft, bedrooms, bathrooms, year_built, lot_size, ...]
sample_input = np.array([[1800, 3, 2, 2005, 8500]])
# Predict
predicted_price = model.predict(sample_input)
print(f"Predicted house price: ${predicted_price[0][0]:,.2f}")
Model Architecture
Input (tabular features)
→ Dense(256, relu) → BatchNormalization → Dropout(0.3)
→ Dense(128, relu) → BatchNormalization → Dropout(0.2)
→ Dense(64, relu)
→ Dense(1, linear) ← regression output
(Update to match your actual architecture)
Limitations & Bias
- Performance is tied to the geographic and temporal distribution of training data — may not generalize to all housing markets
- Does not account for macroeconomic factors (interest rates, market trends) that heavily influence real prices
- Outliers (luxury properties, distressed sales) may be predicted with lower accuracy
Citation
@misc{pasumarthi2026houseprices,
author = {Chandrasekar Adhithya Pasumarthi},
title = {House Price Prediction Model},
year = {2026},
publisher = {Hugging Face},
url = {https://huggingface.co/Chandrasekar123/PredictingHousePrices}
}
Contact
- GitHub: @Adhithpasu