INFORMATION PAPER International Journal of Recent Trends in Engineering, Vol. 1, No. 5, May 2009 172 The Effect of the Foundation Subject to the Final Year Subject K.Kadirgama 1 , M.M.Noor 1 , M.S.M.Sani 1 , Abdullah Adnan Mohamed 2 , Rosli A.Bakar 3 , Abdullah Ibrahim 4 1 Faculty of Mechanical Engineering, Universiti Malaysia Pahang, 26300 UMP, Kuantan, Pahang, MALAYSIA Phone: +609-5492223 Fax: +609-5492244, Email: kumaran@ump.edu.my / muhamad@ump.edu.my 2 Centre for Modern Languages & Human Sciences, University Malaysia Pahang 3 Automotive Excellent Center, Universiti Malaysia Pahang, Email: rosli@ump.edu.my 4 Academic Staff Development Center, Universiti Malaysia Pahang, Email: abi@ump.edu.my Abstract - This paper discussed the foundation or pre-requisite subjects influence the student’s performance in Heat Transfer subject in University Malaysia Pahang (UMP) and its importance to achieve Outcome Based Education (OBE). Thermodynamics I and II are pre-requisite subject for Heat Transfer. Randomly 30 mechanical engineering students were picked to analysis their performance. Regression analysis and Neural Network were used to prove the effect of pre-requisite subject toward Heat transfer. The analysis shows that Thermodynamics I highly affect the performance of Heat transfer and the students who excellent in Thermodynamics I, will achieve excellent performance in Thermodynamics II and Heat transfer. This result shows that excellent foundations on early years will effect to the final year result directly. Keywords: Foundation, Thermodynamics, Heat Transfer, Outcome Based Education I. INTRODUCTION Early year in engineering study, there are few basic subject which are pre-requisite or foundations to next few subject. Example Thermodynamics I is pre-requisite for Thermodynamics II and further in Heat Transfer. Pre- requisite means course required as preparation for entry into a more advanced academic course or program [1]. Regression analysis is a technique used for the modelling and analysis of numerical data consisting of values of a dependent variable (response variable) and of one or more independent variables (explanatory variables). The dependent variable in the regression equation is modelled as a function of the independent variables, corresponding parameters ("constants"), and an error term. The error term is treated as a random variable. It represents unexplained variation in the dependent variable. The parameters are estimated so as to give a "best fit" of the data. Most commonly the best fit is evaluated by using the least squares method, but other criteria have also been used [1]. Regression can be used for prediction (including forecasting of time-series data), inference, hypothesis testing and modelling of causal relationships. These uses of regression rely heavily on the underlying assumptions being satisfied. Regression analysis has been criticized as being misused for these purposes in many cases where the appropriate assumptions cannot be verified to hold [1, 2]. One factor contributing to the misuse of regression is that it can take considerably more skill to critique a model than to fit a model [3]. However, when a sample consists of various groups of individuals such as males and females, or different intervention groups, regression analysis can be performed to examine whether the effects of independent variables on a dependent variable differ across groups, either in terms of intercept or slope. These groups can be considered from different populations (e.g., male population or female population), and the population is considered heterogeneous in that these subpopulations may require different population parameters to adequately capture their characteristics. Since this source of population heterogeneity is based on observed group memberships such as gender, the data can be analyzed using regression models by taking into consideration multiple groups. In the methodology literature, subpopulations that can be identified beforehand are called groups [4, 5].Model can account for all kinds of individual differences. Regression mixture models described here are a part of a general framework of finite mixture models [6] and can be viewed as a combination of the conventional regression model and the classic latent class model [7]. It should be noted that there are various types of regression mixture models [7], but this only focus on the linear regression mixture model. The following sections will first describe some unique characteristics of the linear regression mixture model in comparison to the conventional linear regression model, including integration of covariates into the model. Second, a step-by-step regression mixture analysis of empirical data demonstrates how the linear regression mixture model may be used by incorporating population heterogeneity into the model. Ko et al. [8] have introduced an unsupervised, self-organised neural network combined with an adaptive time-series AR modelling algorithm to monitor tool breakage in milling operations. The machining parameters and average peak force have been used to build the AR model and neural network. Lee and Lee [9] have used a neural network-based approach to show that by using the force ratio, flank wear can be predicted within 8% to 11.9% error and by using force increment; the © 2009 ACADEMY PUBLISHER