Modified LVQ based clustering analysis for decision making in construction management

Abstract

Extracting the data and information from manual data repository is difficult, costly and time-consuming. They have prospects for making decision in construction process. Decision making can be performed by collecting data timely and cost effectively from the data warehouse by providing a model of the decision making process and programming pertinent knowledge into it. Data mining automates the process of finding predictive information from the large databases. To improve the decision making in construction management the artificial neural network (ANN) commonly known as neural network (NN) is one of the method which can be used in Data Mining. Clustering is one of the basic data analysis method used in data mining. In Construction Management, the problem is how to analyze the data to obtain quick analysis for the extraction of useful Clusters. In this research we have applied modified Learning Vector Quantization (LVQ) neural network to classify the construction projects into flexible Clusters for dynamic analysis. These examples are based upon past experiences of similar data, it can identify the problem and suggest suitable alternatives. The proposed modified LVQ technique is efficient with respect to the number of clusters and time. This system is fast and the accuracy of the system has been verified by domain experts through numerous case examples.