Nan-Hsing Chiu

Integration of the grey relational analysis with genetic algorithm for software effort estimation

Accurate estimates of efforts in software development are necessary in project management practices. Project managers or domain experts usually conduct software effort estimation using their experience; hence, subjective or implicit estimates occur frequently. As most software projects have incomplete information and uncertain relations between effort drivers and the required development effort, the grey relational analysis (GRA) method has been applied in building a formal software effort estimation model for this study. The GRA in the grey system theory is a problem-solving method that is used when dealing with similarity measures of complex relations. This paper examines the potentials of the software effort estimation model by integrating a genetic algorithm (GA) to the GRA. The GA method is adopted to find the best fit of weights for each software effort driver in the similarity measures. Experimental results show that the software effort estimation using an integration of the GRA with GA method presents more precise estimates over the results using the case-based reasoning (CBR), classification and regression trees (CART), and artificial neural networks (ANN) methods.

Applying fuzzy neural network to estimate software development effort

Abstract The ability to accurately and consistently estimate software development efforts is required by the project managers in planning and conducting software development activities. Since software effort drivers are vague and uncertain, software effort estimates, especially in the early stages of the development life cycle, are prone to a certain degree of estimation errors. A software effort estimation model which adopts a fuzzy inference method provides a solution to fit the uncertain and vague properties of software effort drivers. The present paper proposes a fuzzy neural network (FNN) approach for embedding artificial neural network into fuzzy inference processes in order to derive the software effort estimates. Artificial neural network is utilized to determine the significant fuzzy rules in fuzzy inference processes. We demonstrated our approach by using the 63 historical project data in the well-known COCOMO model. Empirical results showed that applying FNN for software effort estimates resulted in slightly smaller mean magnitude of relative error (MMRE) and probability of a project having a relative error of less than or equal to 0.25 (Pred(0.25)) as compared with the results obtained by just using artificial neural network and the original model. The proposed model can also provide objective fuzzy effort estimation rule sets by adopting the learning mechanism of the artificial neural network.

A comparative evaluation on the accuracies of software effort estimates from clustered data

Precision in estimating the required software development effort plays a critical factor in the success of software project management. Most existing software effort estimation models only compare the accuracies of software effort estimates from the historical data without clustering. A potential factor that can affect the accuracies of the established effort estimation models is the homogeneity of the data. However, such investigation on the effects of the accuracies of the derived effort estimates is seldom explored in software effort estimation literature. Therefore, this paper aims to explore the effects of accuracies of the software effort estimation models established from the clustered data by using the International Software Benchmarking Standards Group (ISBSG) repository. The ordinary least square (OLS) regression method is adopted to establish a respective effort estimation model in each cluster of datasets. The empirical experiment results show that the estimation accuracies do not reveal significant differences within the respective dataset clustered by each software effort driver. It also demonstrates that software effort estimation models from the clustered data present almost similar accuracy results compared to models from the entire data without clustering.

A case-based classifier for hypertension detection

The exploration of three-dimensional (3D) anthropometry scanning data along with other existing subject medical profiles using data mining techniques becomes an important research issue for medical decision support. This research attempts to construct a classification approach based on the hybrid use of case-based reasoning (CBR) and genetic algorithms (GAs) for hypertension detection using anthropometric body surface scanning data. The obtained result reveals the relationship between a subjects 3D scanning data and hypertension disease. The GA is adopted to determine the appropriate feature weights for CBR. The proposed approaches were experimented and compared with a regular CBR and other widely used approaches including neural nets and decision trees. The experiment showed that applying GA to determine the suitable weights in CBR is a feasible approach to improving the effectiveness of case matching of hypertension disease. It also demonstrated that different weighted CBR approach presents better classification accuracy over the results obtained from other approaches.

Combining techniques for software quality classification: An integrated decision network approach

Accurately predicting fault-prone modules is a major problem in quality control of a software system during software development. Selecting an appropriate suggestion from various software quality classification models is a difficult decision for software project managers. In this paper, an integrated decision network is proposed to combine the well-known software quality classification models in providing the summarized suggestion. A particle swarm optimization algorithm is used to search for suitable combinations among the software quality classification models in the integrated decision network. The experimental results show that the proposed integrated decision network outperforms the independent software quality classification models. It also provides an appropriate summary for decision makers.

An early software-quality classification based on improved grey relational classifier

The inherent uncertainty and incomplete information of the software development process presents particular challenges for identifying fault-prone modules and providing a preferred model early enough in a development cycle in order to guide software enhancement efforts effectively. Grey relational analysis (GRA) of grey system theory is a well known approach that is utilized for generalizing estimates under small sample and uncertain conditions. This paper examines the potential benefits for providing an early software-quality classification based on improved grey relational classifier. The particle swarm optimization (PSO) approach is adopted to explore the best fit of weights on software metrics in the GRA approach for deriving a classifier with preferred balance of misclassification rates. We have demonstrated our approach by using the data from the medical information system dataset. Empirical results show that the proposed approach provides a preferred balance of misclassification rates than the grey relational classifiers without using PSO. It also outperforms the widely used classifiers of classification and regression trees (CART) and C4.5 approaches.

Opinion mining of hotel customer-generated contents in Chinese weblogs

Customer-generated contents in weblogs provide tourism organisations with valuable market intelligence and ongoing market research opportunities. In this study, an opinion mining method based on feature-based sentiment classification is proposed to extract the online electronic word-of-mouth on weblogs in Taiwan. For opinion extraction, a supervised semantic orientation using the point-wise mutual information (SO_PMI) algorithm based on the extension of Turneys unsupervised SO_PMI algorithm is proposed to extract the opinion words. In addition, a heuristic n-phrase rule is proposed to find out customer opinions about hotel attributes, including hotel image, services, price/value, food and beverage, room, amenities, and location. The experimental results show that the proposed approach mixed with supervised SO_PMI algorithm and heuristic n-phrase rule can demonstrate its effectiveness with acceptable classification and forecasting performances. Furthermore, a perceptual map based on correspondence analysis visually presents opinions comparison to provide the insight of competitive positions.

Online detection of concerned HIV-related messages in web forums

Web forums become the means of online communication and information sharing sources for the learning about health care and related treatment knowledge. By adopting web crawlers and natural language processing techniques, the automatic identification approach of the concerned HIV-related messages is proposed to facilitate the health authorities and social support groups in instant counseling. The proposed supervised GA/k-means for classification approach can help construct an effective identification and classification model with acceptable classification performance accompanied with its full flexibility to develop different fitness functions in accordance with the need of different requirements. Furthermore, with the aid of correspondence analysis, the most frequently used terms in concerned HIV-related messages are identified and focus on risky sexual behavior whereas unconcerned messages are those who of worried well.

Reconstruction of Banknote Fragments Based on Keypoint Matching Method

Banknotes may be shredded by a scrap machine, ripped up by hand, or damaged in accidents. This study proposes an image registration method for reconstruction of multiple sheets of banknotes. The proposed method first constructs different scale spaces to identify keypoints in the underlying banknote fragments. Next, the features of those keypoints are extracted to represent their local patterns around keypoints. Then, similarity is computed to find the keypoint pairs between the fragment and the reference banknote. The banknote fragments can determine the coordinate and amend the orientation. Finally, an assembly strategy is proposed to piece multiple sheets of banknote fragments together. Experimental results show that the proposed method causes, on average, a deviation of 0.12457 ± 0.12810° for each fragment while the SIFT method deviates 1.16893 ± 2.35254° on average. The proposed method not only reconstructs the banknotes but also decreases the computing cost. Furthermore, the proposed method can estimate relatively precisely the orientation of the banknote fragments to assemble.

Application of particle swarm optimization for improving the identification of image objects

Flight safety is very important issue for aviation industries. Analyzing the flight accidents on the basis of 2-dimensional image is hardly to illustrate the complex injuries of passengers in the flight cabin. However, how to illustrate the flight accident is a challenge from 2-dimensional space to 3-dimensional space. This study proposes a particle swarm optimization approach for improving the identification of objects from 2-dimensional image. The recognition results provide the information for building 3-dimensional systems for flight accident investigators. The experiments also show that it is a feasible approach for improving the identification of image objects.