Search results

The purpose of this paper is to apply Poisson generalized linear model (PGLM) with log link instead of multinomial logistic regression to monitor multinomial logistic profiles in Phase I. Hence, estimating the coefficients becomes easier and more accurate.

Simulation technique is used to assess the performance of the proposed algorithm using four different control charts for monitoring.

The proposed algorithm is faster and more accurate than the previous algorithms. Simulation results also indicate that the likelihood ratio test method is able to detect out-of-control parameters more efficiently.

The PGLM with log link has not been used to monitor multinomial profiles in Phase I.

This study aims to investigate the effects of process input parameters (welding current, welding time, electrode pressure and holding time) on the output responses (nugget diameter, peak load and indentation) that control the mechanical properties and quality of the joints in dissimilar resistance spot welding (RSW) for the third generation of advanced high-strength steel (AHSS) quenching and partitioning (Q&P980) and (SPFC780Y) high-strength steel spot welds.

Design of experiment approach with two level factors and center points was adopted. Destructive peel and shear tensile strengths were used to measure the responses. The significant factors were determined using analysis of variance implemented by Minitab 18 software. Finally, multiresponse optimization was carried out using the desirability function analysis method.

Holding time was the most significant factor influencing nugget diameter, whereas welding current had the greatest impact on peak load and indentation. Multiresponse optimization revealed that the optimal settings were a welding current of 12.5 KA, welding time of 18 cycles, electrode pressure of 420 Kgf and holding time of 10 cycles. These settings produced a nugget diameter of 8.0 mm, a peak load of 35.15 KN and an indentation of 22.5%, with a composite desirability function of 0.764.

This study provides an effective approach for multiple response optimization to the mechanical behavior of RSW joints, even though there have been few studies on the third generation of AHSS joints and none on the dissimilar joints of the materials used in this study.

A crucial decision in financial services is how to classify credit or loan applicants into good and bad applicants. The purpose of this paper is to propose a four-stage hybrid data mining approach to support the decision-making process.

The approach is inspired by the bagging ensemble learning method and proposes a new voting method, namely two-level majority voting in the last stage. First some training subsets are generated. Then some different base classifiers are tuned and afterward some ensemble methods are applied to strengthen tuned classifiers. Finally, two-level majority voting schemes help the approach to achieve more accuracy.

A comparison of results shows the proposed model outperforms powerful single classifiers such as multilayer perceptron (MLP), support vector machine, logistic regression (LR). In addition, it is more accurate than ensemble learning methods such as bagging-LR or rotation forest (RF)-MLP. The model outperforms single classifiers in terms of type I and II errors; it is close to some ensemble approaches such as bagging-LR and RF-MLP but fails to outperform them in terms of type I and II errors. Moreover, majority voting in the final stage provides more reliable results.

The study concludes the approach would be beneficial for banks, credit card companies and other credit provider organisations.

A novel four stages hybrid approach inspired by bagging ensemble method proposed. Moreover the two-level majority voting in two different schemes in the last stage provides more accuracy. An integrated evaluation criterion for classification errors provides an enhanced insight for error comparisons.

The purpose of this paper is to compare the performances of np-VP control chart with estimated parameter to the np-VP control chart with known parameter using average time-to-signal (ATS), standard deviation of the time-to-signal (SDTS), and average number of observations to signal (ANOS) as performance measures.

The approach used in this study is probabilistic in which the expected values of performance measures are calculated using probabilities of different estimators used to estimate process parameter.

Numerical results indicate different performances for the np-VP control chart in known and estimated parameter cases. It is obvious that when process parameter is not known and is estimated using Phase I data, the chart does not perform as user expects. To tackle this issue, optimal Phase I estimation scenarios are recommended to obtain the best performance from the chart in the parameter estimation case in terms of performance measures.

This research adds to the body of knowledge in quality control of process monitoring systems. This paper may be of particular interest to practitioners of quality systems in factories where products are monitored to reduce the number of defectives and np chart parameter needs to be estimated.

The originality of this paper lies within the context in which an adaptive np control chart is studied and the process parameter unlike previous studies is assumed unknown. Although other types of control charts have been studied when process parameter is unknown but this is the first time that adaptive np chart performance with estimated process parameter is studied.