Real estate valuation is often complex and subjective, making it difficult for buyers and sellers to determine the true value of a property. In this project, we analyze the Ames Housing Dataset, which contains 1,460 observations and 81 features, to provide data-driven insights into the housing market. Our study focuses on two core business objectives: Regression Task: Can we accurately predict the exact SalePrice of a house based on its attributes? Classification Task: Can we distinguish between "High Value" (above median price) and "Low Value" properties? We followed a structured data science pipeline (CRISP-DM), moving from rigorous Data Cleaning and domain-specific Feature Engineering to training advanced machine learning models, including Random Forest and Logistic Regression. This report details our methodology, model comparisons, and key findings regarding the drivers of house prices.