An ML-based Intelligent System for House Cost Estimation and Space Optimization
Divvela Srinivasa Rao, Kiranmai Merum, M.Srikanth, M.Lakshmi Narayana, M.Chandra Naik, Ch Lalitha Syama Sundari
Abstract
The main aim of this study is to come up with a smart machine learning driven system that can provide a good estimate of the costs involved in constructing a house and predict the most efficient space use under the stipulated parameters that are provided by the user like the area of land, budget, and orientation. The proposed system combines predictive modeling and data-driven learning to overcome the drawbacks of the conventional traditional approach of using simple estimation, which typically does not consider interdependent design and financial factors. The two machine learning models, a gradient boosting regressor (GBR) to estimate the cost of the house and a support vector regressor (SVR) to optimize the spatial, were trained on a dataset of 500 completed projects of houses. Each of the models was compared based on the mean absolute error (MAE) and the R 2-score to determine the level of estimation. The experimental results demonstrate that the proposed system has high reliability, scalability and flexibility in its cost and utilization efficiency estimation. It has been established that increased plot sizes and correct orientations result in optimized space and cost estimation. The study offers a decision-support process that facilitates planning of construction, efficiency of resources, as well as smarter management to open the path to data-driven and intelligent construction design.