Abstract:
PURPOSE: A CBR-based cost prediction system for the pre-design and schematic design phase of public multi-housing construction projects, and a model are provided to improve the accuracy and reliability in estimation by developing a public apartment construction expense predictive model based on a construction expenses per unit extracted by a CBR(Case Based Reasoning) algorithm. CONSTITUTION: An input unit(10) inputs apartment information of a target project, and a feature weight calculator(21) calculates the feature weight for the construction cost influence factor. A feature similarity calculator(22) calculates the feature similarity based on the calculated feature weight. A unit cost calculator(24) calculates the construction coast per unit area of the extracted similarity sample.