Real Estate Valuation with TensorFlow

This notebook presents a real estate valuator model built to predict the price per unit area of properties in Sindian Dist., New Taipei City, Taiwan. Utilizing a dataset from the University of California Irvine containing historical transaction data from 2012-2013, the model employs linear regression techniques. The dataset comprises 414 instances with features like transaction date, house age, distance to the nearest MRT station, number of convenience stores, and geographical coordinates. The target variable represents the house price per unit area, denominated in 10,000 New Taiwan Dollar/Ping. The methodology encompasses data acquisition from the University of California Irvine, preprocessing where data is split into training and testing sets and standardized, model building and training using TensorFlow, and evaluation. After building and training the model, users can predict prices for new properties by entering the property's details into the notebook.