Forecasting techniques suitable to estimate unitary charges price indexes of PFI data: Context of northern region Peninsular Malaysia

摘要

The Private Financial Initiative (PFI) program has been implemented by many developed countries throughout the world as an innovative way for the governments to improve future public service delivery and infrastructure procurement. However, the idea is just about to germinate in Malaysia and its success is still vague. The major phase that needs to be given main attention in this agenda is value for money whereby optimum efficiency and effectiveness of each expense is attained. Therefore, at the early stage of this study, estimating unitary charges or materials price indexes in each region in Malaysia was the key objective. In this paper, we aim to discover the best forecasting method to estimate unitary charges price indexes in construction industry by different states in the northern region of Peninsular Malaysia (Pulau Pinang, Kedah and Perlis). The unitary charges indexes data used were monthly data from year 2002 to 2011 of different states in the northern region Peninsular Malaysia, comprise of price indexes of aggregate, sand, steel reinforcement, ready mix concrete, bricks and partition, roof material, floor and wall finishes, ceiling, plumbing materials, sanitary fittings, paint, glass, steel and metal sections, timber and plywood. At the end of the study, it was found that backpropagation neural network with linear transfer function produced the most accurate and reliable results for estimating unitary charges price indexes in every states in northern region Peninsular Malaysia based on the Root Mean Squared Errors, where the values for both estimation and evaluation sets were approximately zero and highly significant at p < 0.01. Therefore, artificial neural network is sufficient to forecast construction materials price indexes in Malaysia. The estimated price indexes of construction materials will contribute significantly to the value for money of PFI as well as towards Malaysian economical growth.