Predicting defect resolution time using cosine similarity

Baloch

Mehran Univ. res. j. eng. technol.

et al. 2021

The development of software undergoes multiple regression phases to deliver quality software. Therefore, to minimize the development effort, time and cost it is very important to understand the probable defects associated with the designed modules. It is possible that occurrence of a range of defects may impact the designed modules which need to be predicted in advance to have a close inter-association with the depended modules. Most of the existing defect prediction classifier mechanisms are derived from the past project data learning, but it is not sufficient for new project defect predicting as the new design may have a different kind of parameters and constraints. This paper recommends Regression Analysis (RA) based defect learning and prediction Defect Prediction (RA-DP) mechanism to support the defective or non-defective prediction for quality software development. The RA-DP approach provides two methods to perform this prediction analysis. It initially presents an association learning through RA to construct the regression rules from the learned knowledge required for the defect prediction. The constructed regression rules are used for defect prediction and analysis. To measure the performance of the RA-DP a regression experimental evaluation is performed over the defect-prone PROMISE dataset from NASA project. The outcome of the results is analyzed through measuring the prediction Accuracy, Sensitivity and Specificity to demonstrate the improvisation and effectiveness of the proposal in comparison to a few existing classifiers.

Section: Background Studymentioning

confidence: 99%

“…To evaluate the proposed prediction mechanism it implements the method in a WEKA Tool environment [21]. It utilizes the data repository of "NASA -Metric Data Program" to perform the comparative analysis for the enhancement.…”

Section: Experiments Evaluationmentioning

confidence: 99%

A Regression Analysis Based Model for Defect Learning and Prediction in Software Development

Baloch

Mehran Univ. res. j. eng. technol.

et al. 2021

“…In the field of software engineering, it is one of the active research focus [11]. Since DEF-PR models include a group of "defect-prone software artefacts" [12], "quality assurance teams" which efficiently assign inadequate resources to test and examine software products [13,14]. Several DEF-PR studies [15][16][17] have been carried out and all are based on an ML approach or a statistical approach.…”

Section: Background Workmentioning

confidence: 99%

“…Experiments were performed utilizing the algorithm implemented in the "WEKA environment" [14] utilizing the "NASA -Metric Data Program (MDP) Repository". Below we discuss data sets, evaluation measurements, and analysis of results.…”

Section: Experimental Evaluationmentioning

confidence: 99%

A Software Defect Learning and Analysis Utilizing Regression Method for Quality Software Development

Jadhav¹

2019

IJATCSE

The program is a complex object consisting of different units with variable degrees of defects. By predicting the effectiveness and frequency of program defects, program managers can make better use of manpower, cost, and time to obtain better quality assurance. It is always possible to have a set of defects that affect designed and predictable units in order to have close association with the subsidiaries. Most of the current defect prediction rating mechanism is derived from learning the previous project data, but it is not sufficient to predict the defect of the new project because the new design may contain a different type of parameter. This paper proposes a Software Defect Learning and Analysis utilizing Regression Method (SDLA-RM) to detect defects and plan a better maintenance strategy, which can support the prediction of a defective or nondefective software unit prior to deployment in any project programs. The SDL-RM mechanism extends Regression Analysis (RAM) to create an effective rulebased model for accurately classifying program faults. This approach improves the predictability of software defects, allowing software development to spend more time testing components that are expected to contain errors. The experimental evaluation is carried out across the NASA-PROMISE repository data sets, that outcome of the results in comparison with existing classifiers suggest the effectiveness and practical perspective in the software development.

“…where • denotes the Euclidean distance, x t and x v are vector spaces with a fixed Cartesian coordinate system, and Σ −1 is the inverse matrix of the group covariances of the class [43][44][45].…”

Section: Similarity Measuresmentioning

confidence: 99%

Multilinear EigenECGs and FisherECGs for Individual Identification from Information Obtained by an Electrocardiogram Sensor

et al. 2018

In this study, we present a third-order tensor-based multilinear eigenECG (MEECG) and multilinear Fisher ECG (MFECG) for individual identification based on the information obtained by an electrocardiogram (ECG) sensor. MEECG and MFECG are based on multilinear principal component analysis (MPCA) and multilinear linear discriminant analysis (MLDA) in the field of multilinear subspace learning (MSL), respectively. MSL directly extracts features without the vectorization of input data, while MSL extracts features without vectorizing the input data while maintaining most of the correlations shown in the original structure. In contrast with unsupervised linear subspace learning (LSL) techniques such as PCA (Principal Component Analysis) and LDA (Linear Discriminant Analysis), it is less susceptible to small-data problems because it learns more compact and potentially useful representations, and it can efficiently handle large tensors. Here, the third-order tensor is formed by reordering the one-dimensional ECG signal into a two-dimensional matrix, considering the time frame. The MSL consists of four steps. The first step is preprocessing, in which input samples are centered. The second step is initialization, in which eigen decomposition is performed and the most significant eigenvectors are selected. The third step is local optimization, in which input data is applied by eigenvectors from the second step, and new eigenvectors are calculated using the applied input data. The final step is projection, in which the resultant feature tensors after projection are obtained. The experiments are performed on two databases for performance evaluation. The Physikalisch-Technische Bundesanstalt (PTB)-ECG is a well-known database, and Chosun University (CU)-ECG is directly built for this study using the developed ECG sensor. The experimental results revealed that the tensor-based MEECG and MFECG showed good identification performance in comparison to PCA and LDA of LSL.