Lee-Ing Tong’s research while affiliated with National Yang Ming Chiao Tung University and other places

Artificial intelligence algorithms and big data analysis methods are commonly employed in network intrusion detection systems. However, challenges such as unbalanced data and unknown network intrusion modes can influence the effectiveness of these methods. Moreover, the information personnel of most enterprises lack specialized knowledge of information security. Thus, a simple and effective model for detecting abnormal behaviors may be more practical for information personnel than attempting to identify network intrusion modes. This study develops a network intrusion detection model by integrating weighted principal component analysis into an exponentially weighted moving average control chart. The proposed method assists information personnel in easily determining whether a network intrusion event has occurred. The effectiveness of the proposed method was validated using simulated examples.

In the modern highly competitive marketplace and global market environment, product quality improvements that abridge development time and reduce the production costs are effective methods for promoting the business competitiveness of a product in shorter lifecycles. Since the design process is the best time to control such parameters, systematically designing the processes to develop a product that more closely fits the demand requirements for the market is a key factor for developing a successful product. In this paper, a combined affective engineering method and grey relational analysis are used to develop a product design process. First, design image scale technology is used to acquire the best the design criteria factors, and then affective engineering methods are used to set the relationships between customer needs and production factors. Finally, grey relational analysis is used to select the optimal design strategy. Using this systematic design method, a higher quality product can be expanded upon in a shorter lead-time for improving business competition.

Constructing a classification model for the multi-class data is a critical problem in many areas. In practical applications, data in multiple classes are often imbalanced which might result in a classification model with high overall accuracy rate but with low accuracy rate for the minority class. However, minority class is usually the more important one compared to other classes in practice. This study integrates dual response surface methodology, logistic regression analysis, and desirability function to develop an optimal re-sampling strategy for classifying multi-class imbalanced data to effectively improve the low classification accuracy rate of the minority class(es) while still maintain a certain accuracy rate for the majority class(es). Three data-sets drawn from KEEL Database were used in the numerical experiments. The results showed that the proposed method can effectively improve the low classification accuracy rate of the minority class in contrast to the previous work.

Mitigating global warming problems initially involves reducing greenhouse gas (GHG) emissions, therefore the uncertainty of GHG emission estimates needs to be assessed concisely. Although the uncertainty of GHG emission estimates is generally evaluated using classical confidence interval, quantifying the uncertainty based on non-normal GHG emission estimates or small dataset may lead to a significant bias. Using bootstrap confidence intervals is an effective means of reducing such a bias. This study presents a procedure for constructing four bootstrap confidence intervals to assess the uncertainty of GHG emission estimates for three non-normal distributions (namely, Weibull, Gamma and Beta). These bootstrap confidence intervals are standard bootstrap (SB) confidence interval, percentile bootstrap (PB) confidence interval, Bias-corrected percentiles bootstrap (BCPB) confidence interval and bias-corrected and accelerated (BCa) confidence interval. The sensitivity of bootstrap intervals for emission data is examined under various combinations of sample size and parameters of underlying non-normal distributions using three indices: coverage performance, interval mean, and interval standard deviation. Simulation results indicate that the bootstrap confidence intervals for assessing the uncertainty of emission estimates are highly applicable with small sample size and the data distribution is non-normal. Compared with the classical confidence interval, bootstrap confidence intervals have smaller interval mean and smaller interval standard deviation for small sample size under non-normal distributions. This study recommends BCa confidence interval to assess the uncertainty of the emission estimates as long as sample size is 15 or more and the distribution is non-normal. A case study with emission data of the High-Bleed Pneumatic controllers demonstrates the effectiveness of the proposed procedure

Traditional credit risk assessment models do not consider the time factor; they only think of whether a customer will default but not the when to default. The result can not provide a manager to make the profit maximum decision. Actually, even if a customer defaults, the financial institution still can gain profit in some conditions. Nowadays, most research applied the Cox proportional hazards (PH) model into their credit scoring models, predicting the time when a customer is most likely to default, to solve the credit risk assessment problem. However, in order to fully utilize the fully dynamic capability of the Cox PH model, time-varying macroeconomic variables are required which involves more advanced data collection. Since short-term default cases are the ones that bring a great loss for a financial institution, instead of predicting when a loan will default, a loan manager is more interested in identifying those applications which may default within a short period of time when approving loan applications. This paper proposed a decision tree-based short-term default credit risk assessment model to assess the credit risk. The goal is to use the decision tree to filter the short-term default to produce a highly accurate model that could distinguish default lending. This paper integrates bootstrap aggregating (Bagging) with a synthetic minority over-sampling technique (SMOTE) into the credit risk model to improve the decision tree stability and its performance on unbalanced data, respectively. Finally, a real case of small and medium enterprise loan data that have been drawn from a local financial institution located in Taiwan is presented to further illustrate the proposed approach. After comparing the result that was obtained from the proposed approach with the logistic regression and Cox proportional hazards models, it was found that the classifying recall rate and precision rate of the proposed model were obviously superior to the logistic regression and Cox proportional hazards models.

Techniques for statistical process control (SPC), such as using a control chart, have recently garnered considerable attention in the software industry. These techniques are applied to manage a project quantitatively and meet established quality and process-performance objectives. Although many studies have demonstrated the benefits of using a control chart to monitor software development processes (SDPs), some controversy exists regarding the suitability of employing conventional control charts to monitor SDPs. One major problem is that conventional control charts require a large amount of data from a homogeneous source of variation when constructing valid control limits. However, a large dataset is typically unavailable for SDPs. Aggregating data from projects with similar attributes to acquire the required number of observations may lead to wide control limits due to mixed multiple common causes when applying a conventional control chart. To overcome these problems, this study utilizes a Q chart for short-run manufacturing processes as an alternative technique for monitoring SDPs. The Q chart, which has early detection capability, real-time charting, and fixed control limits, allows software practitioners to monitor process performance using a small amount of data in early SDP stages. To assess the performance of the Q chart for monitoring SDPs, three examples are utilized to demonstrate Q chart effectiveness. Some recommendations for practical use of Q charts for SDPs are provided.

Solar energy has become an important energy source in recent years as it generates less pollution than other energies. A photovoltaic (PV) system, which typically has many components, converts solar energy into electrical energy. With the development of advanced engineering technologies, the transfer efficiency of a PV system has been increased from low to high. The combination of components in a PV system influences its transfer efficiency. Therefore, when predicting the transfer efficiency of a PV system, one must consider the relationship among system components. This work accurately predicts whether transfer efficiency of a PV system is high or low using a novel hybrid model that combines rough set theory (RST), data envelopment analysis (DEA), and genetic programming (GP). Finally, real data-set are utilized to demonstrate the accuracy of the proposed method.

In traditional scheduling problems, the processing time for the given job is assumed to be a constant regardless of whether the job is scheduled earlier or later. However, the phenomenon named “learning effect” has extensively been studied recently, in which job processing times decline as workers gain more experience. This paper discusses a bi-criteria scheduling problem in an m-machine permutation flowshop environment with varied learning effects on different machines. The objective of this paper is to minimize the weighted sum of the total completion time and the makespan. A dominance criterion and a lower bound are proposed to accelerate the branch-and-bound algorithm for deriving the optimal solution. In addition, the near-optimal solutions are derived by adapting two well-known heuristic algorithms. The computational experiments reveal that the proposed branch-and-bound algorithm can effectively deal with problems with up to 16 jobs, and the proposed heuristic algorithms can yield accurate near-optimal solutions.