ESTATE MANAGEMENT PROJECT TOPICS AND MATERIALS ON EXPERT SYSTEM-BASED PREDICTIVE COST MODEL FOR BUILDING WORKS: NEURAL NETWORK APPROACH.
Cost is one of the three main challenges for the construction manager, where the success of a project is judged by meeting the criteria of cost with budget, schedule on time, and quality as specified by the owner (Rezaian, 2011). In which, poor strategy or incorrect budget or schedule forecasting can easily turn an expected profit into loss (Cheng, et al., 2010). Therefore, effective estimating is one of the main factors of a construction project success (Al-Shanti, 2003). Accordingly, cost estimate in early stage plays a significant role in any construction project (Ayed, 1997), where it allows owners and planners to evaluate project feasibility and control costs effectively (Feng, et al., 2010).In addition, the cost of a building is significantly affected by decisions made at the early phase. While this influence decreases through all phases of building project (Gunaydın & Dogan, 2004).
Due to this prominence of cost estimate in early stage and limited availability of information during the early phase of a project, construction managers typically leverage their knowledge, experience and standard estimators to estimate project costs. As such, intuition plays a significant role in decision-making. Inasmuch the essential needs of project owners and planners to a tool to help them in their early decisions; researchers have worked hard to develop cost estimate technique that maximize the practical value of limited information in order to improve the accuracy and reliability of cost estimation work (Cheng, et al., 2010). Thus, many methods either traditional or artificial intelligence methods were studied and examined for their validity in estimating the project cost at conceptual stage.
In the last years a new approach, based on the theory of computer systems that simulate the learning effect of the human brain as Artificial Neural Networks (ANNs) has grown in popularity (Cavalieri, et al., 2004). One major benefit of using ANN is its ability to understand and simulate more complex functions than older methods such as linear regression (Weckman, et al., 2010). In addition, it can approximate functions well without explaining them. This means that an output is generated based on different input signals and by training those networks, accurate estimates can be generated. (Verlinden, et al., 2007).