The term”regression” was devised by Francis Galton in the nineteenth century todescribe a biological phenomenon. Regression analysis is a statistical methodfor exploring the correlations between independent (xi) and dependent (y)variables. Its purpose is to find the equation that most accurately predictsthe dependent variable as a linear function of two or more independentvariables according to the causal relationships among changes in internalfactors. Regression analysis includes many techniques for modeling andanalyzing several variables, when the focus is on the relationship between adependent variable and one or more independent variables (or ‘predictors’) Regression analysisalso allows to assess the effects of variables measured on different scales,these benefits help researchers / data analysts / data scientists to eliminateand evaluate the best set of variables to be used for building predictivemodels.

1.1.1 Artificial Neural Network (ANN)Artificial neuralnetworks are purely data driven models which through training iterativelytransition from a random state to a final model. They do not depend on expectations aboutfunctional form, probability distribution or smoothness, ANN is a branch ofartificial intelligence (AI) in which structure is based on the biologicalnervous system. It can exhibit a surprising number of the human brain’scharacteristics, e.g. learn from experience and generalize from previousexamples to new problems.

ANN can provide expressive answers even when the datato be processed include errors or are incomplete, and can process informationextremely rapidly when applied to solve real world problems (Lippmann 1988;Smith 1993). Neural Networks (NN) are used in construction for modeling complexrelationships between inputs and outputs or to find patterns in data. An ANN is typicallydefined by three types of parameters (Nitin kumara et al 2013): 1-The interconnection patternbetween different layers of neurons. 2-The learning process for updatingthe weights of the interconnections. 3-The activation function thatconverts a neuron’s weighted input to its output activation. Garrett (1994) hasgiven an interesting engineering definition of the ANN as: “a computationalmechanism able to acquire, represent, and compute mapping from one multivariatespace of information to another, given a set of data representing thatmapping.

” Artificial neural networks (ANNs) are a functional abstraction of thebiologic neural structures of the central nervous system Zayed and Halpin (2005)defined ANN as “the process of developing the ability to generalize, whichcorrectly classify new patterns or to make forecasts and predictions”. earning andgeneralization of the problems in the ANN models could be achieved in spite ofincomplete or erroneous data. The data used to test the prediction capabilityof the network is selected from the total data set. After a predeterminednumber of iterations, the network training and testing is stopped whereas noamelioration is done in the output. ANN has some advantagesincludes (Nitin kumara et al 2013): · Adaptive learning: An ability tolearn how to do tasks based on the data given for training or initialexperience. · Self-Organization: An ANN can create its own organization orrepresentation of the information it receives during learning time. Also, ANN has some disadvantages include(Nitin kumara et al 2013): · ANN can be used only if trainingdata is available. It is not necessary to have a mathematical model of theproblem of interest, and there is no need to provide any form of priorknowledge.

· The solution obtained from thelearning process usually cannot be interpreted. · Most neural network architecturesare black boxes. They cannot be checked whether their solution is probable,i.e. their final state cannot be interpreted in terms of rules. Applications of ANNs inconstruction management go back to the early 1990s. These applications covermany critical topics in productivity, quality, time estimation, bidding,performance evaluation.

Among the numerousartificial neural networks that have been proposed, backpropagation networkshave been extremely popular for their unique learning capability. 80% ofpractical ANN applications used the backpropagation neural networks (Haykin1999). Artificial neuralnetworks (ANNs) models may be called by different names: (1) connectionist models; (2) paralleldistributed processing models; (3) neuromorphicsystems; and (4) neural computing.