Missing data imputation based on low-rank recovery and semi-supervised regression for software effort estimation

Jing, Xiao‐Yuan; Qi, Fumin; Wu, Fei; Xu, Bin

doi:10.1145/2884781.2884827

Cited by 23 publications

(9 citation statements)

References 56 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this section, we deal with the explication of the proposed method. First, the independent variables are normalised by mapping to [0, 1] interval using (1).…”

Section: The Proposed Methodsmentioning

confidence: 99%

See 1 more Smart Citation

A Dataset-Independent Model for Estimating Software Development Effort Using Soft Computing Techniques

Khazaiepoor

Bardsiri

Keynia

2019

Applied Computer Systems

View full text Add to dashboard Cite

During the recent years, numerous endeavours have been made in the area of software development effort estimation for calculating the software costs in the preliminary development stages. These studies have resulted in the offering of a great many of the models. Despite the large deal of efforts, the substantial problems of the offered methods are their dependency on the used data collection and, sometimes, their lack of appropriate efficiency. The current article attempts to present a model for software development effort estimation through making use of evolutionary algorithms and neural networks. The distinctive characteristic of this model is its lack of dependency on the collection of data used as well as its high efficiency. To evaluate the proposed model, six different data collections have been used in the area of software effort estimation. The reason for the application of several data collections is related to the investigation of the model performance independence of the data collection used. The evaluation scales have been MMRE, MdMRE and PRED (0.25). The results have indicated that the proposed model, besides delivering high efficiency in contrast to its counterparts, produces the best responses for all of the used data collections.

show abstract

“…In this section, we deal with the explication of the proposed method. First, the independent variables are normalised by mapping to [0, 1] interval using (1).…”

Section: The Proposed Methodsmentioning

confidence: 99%

“…Software development effort estimation is regarded as an important stage in software development projects. Therefore, many industry specialists and researchers have been devoting their attention to this field during recent years [1].…”

Section: Introductionmentioning

confidence: 99%

A Dataset-Independent Model for Estimating Software Development Effort Using Soft Computing Techniques

Khazaiepoor

Bardsiri

Keynia

2019

Applied Computer Systems

View full text Add to dashboard Cite

show abstract

“…Fang et al [36, 37] proposed a method combined the semi‐supervised clustering learning with the non‐negative LRR framework, which achieved acceptable clustering performance and was robust to different types of noises. Jing et al [38] used the low‐rank recovery and the semi‐supervised regression technique to perform the missing data imputation for software effort estimation.…”

Section: Related Workmentioning

confidence: 99%

Low‐rank representation for semi‐supervised software defect prediction

Zhang

Jing

2018

IET softw.

Self Cite

View full text Add to dashboard Cite

Software defect prediction based on machine learning is an active research topic in the field of software engineering. The historical defect data in software repositories may contain noises because automatic defect collection is based on modified logs and defect reports. When the previous defect labels of modules are limited, predicting the defect-prone modules becomes a challenging problem. In this study, the authors propose a graph-based semi-supervised defect prediction approach to solve the problems of insufficient labelled data and noisy data. Graph-based semi-supervised learning methods used the labelled and unlabelled data simultaneously and consider them as the nodes of the graph at the training phase. Therefore, they solve the problem of insufficient labelled samples. To improve the stability of noisy defect data, a powerful clustering method, low-rank representation (LRR), and neighbourhood distance are used to construct the relationship graph of samples. Therefore, they propose a new semi-supervised defect prediction approach, named low-rank representation-based semi-supervised software defect prediction (LRRSSDP). The widely used datasets from NASA projects and noisy datasets are employed as test data to evaluate the performance. Experimental results show that (i) LRRSSDP outperforms several representative state-of-the-art semi-supervised defect prediction methods; and (ii) LRRSSDP can maintain robustness in noisy environments.

show abstract

“…In the last couple of years, there have been major advancements in the domain of missing data imputation. Various techniques include amongst others: Bayesian methods [19,13,1], Nearest Neighbours, Mean and Mode, Random Forests [18], Latent Class Models [21], Multiple Correspondence Analysis [2] Hybrid approaches [17,3,9] and Neural Networks [10] more recently. Different methods have variable performance based on the missing data mechanism and the structure of the data.…”

Section: Related Workmentioning

confidence: 99%