• G.C.Sarode
• Shubham E.Chandgude

Cost estimation is an experience-based task, which involves evaluations of unknown circumstances and complex relationships of cost-influencing factors. An artificial neural network (ANN) is an analogy-based process, which best suits the cost forecasting domain. The primary advantages of ANNs include their ability to learn by examples (past projects), and to generalize solutions for forthcoming applications (future projects). Construction cost prediction is important for construction firms to compete and grow in the industry. Accurate construction cost prediction in the early stage of project is important for project feasibility studies and successful completion. There are many factors that affect the cost prediction. This research presents the comparison between ANN and SVM. The objective of this paper is to develop neural networks and multilayer perceptron based model for construction cost prediction. Estimating construction costs and predicting price escalation are major steps for project owners, estimators, and contractors. Therefore, the cost estimation plays a significant role in construction project decisions and represents the most important corner in iron triangle of construction management. This project is implemented on actual site case study located at Ambegaon Pune under the management of SP Construction Pune (Ganesh Construction).In order to success theconstruction projects we need technique to estimate the cost with high degree of accuracy and less error. In the present investigation, the model built by applying both quantitative approach and qualitative approach to identify the factors (variables) as inputs of the model. 85 projects are used for developing, training and testing the ANNs model, the output of this model is the expected construction costs of the projects. To validate the model, 14 projects as sample have tested to predict the cost with high degree accuracy and acceptable error. Consumer Price Index, cost of construction materials, type of building, market conditions, structural system, site Area, type of slab, other Supplementary buildings, location of the Project, project Size, type of foundation, building closeness, and fluctuation in the Currency are the main factors affecting in construction buildings costs. These factors have been used as inputs in ANN model and all data is extracted from the historical projects, the model has been developed and trained for 70 projects and compared the actual cost with predicted cost. The model was validated throughout sample of projects. 13- 17- 1 model was the best between 15 models are developed, 6% is the mean absolute percentage error for model is tested. The results are clearly provided a good indicator for predicting the construction buildings costs in the future with high degree of accuracy.

Cost Factors, Artificial Neural Network System, Feed Forward Network,Multilayer Perceptron, Back Propagation Error, Developing the model.

"TO STUDY COST PREDICTION ANALYSIS OF CONSTRUCTION PROJECT USING ANN MODEL (ARTIFICIAL NEURAL NETWORK) and SVM BY MATLAB", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.7, Issue 7, page no.291-306, July-2020, Available :http://www.jetir.org/papers/JETIR2007037.pdf

"TO STUDY COST PREDICTION ANALYSIS OF CONSTRUCTION PROJECT USING ANN MODEL (ARTIFICIAL NEURAL NETWORK) and SVM BY MATLAB", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.7, Issue 7, page no. pp291-306, July-2020, Available at : http://www.jetir.org/papers/JETIR2007037.pdf