Data Mining Practical Machine Learning Tools and Techniques

Context Systematic mapping studies are used to structure a research area, while systematic reviews are focused on gathering and synthesizing evidence. The most recent guidelines for systematic mapping are from 2008. Since that time, many suggestions have been made of how to improve systematic literature reviews (SLRs). There is a need to evaluate how researchers conduct the process of systematic mapping and identify how the guidelines should be updated based on the lessons learned from the existing systematic maps and SLR guidelines. Objective To identify how the systematic mapping process is conducted (including search, study selection, analysis and presentation of data, etc.); to identify improvement potentials in conducting the systematic mapping process and updating the guidelines accordingly. Method We conducted a systematic mapping study of systematic maps, considering some practices of systematic review guidelines as well (in particular in relation to defining the search and to conduct a quality assessment). Results In a large number of studies multiple guidelines are used and combined, which leads to different ways in conducting mapping studies. The reason for combining guidelines was that they differed in the recommendations given. Conclusion The most frequently followed guidelines are not sufficient alone. Hence, there was a need to provide an update of how to conduct systematic mapping studies. New guidelines have been proposed consolidating existing findings.

Guidelines for conducting systematic mapping studies in software engineering : An update

Background: The accurate prediction of where faults are likely to occur in code can help direct test effort, reduce costs, and improve the quality of software. Objective: We investigate how the context of models, the independent variables used, and the modeling techniques applied influence the performance of fault prediction models. Method: We used a systematic literature review to identify 208 fault prediction studies published from January 2000 to December 2010. We synthesize the quantitative and qualitative results of 36 studies which report sufficient contextual and methodological information according to the criteria we develop and apply. Results: The models that perform well tend to be based on simple modeling techniques such as Naive Bayes or Logistic Regression. Combinations of independent variables have been used by models that perform well. Feature selection has been applied to these combinations when models are performing particularly well. Conclusion: The methodology used to build models seems to be influential to predictive performance. Although there are a set of fault prediction studies in which confidence is possible, more studies are needed that use a reliable methodology and which report their context, methodology, and performance comprehensively.

/pdf/a-systematic-literature-review-on-fault-prediction-50xbihfieh.pdf

A Systematic Literature Review on Fault Prediction Performance in Software Engineering

Context: We are strong advocates of evidence-based software engineering (EBSE) in general and systematic literature reviews (SLRs) in particular. We believe it is essential that the SLR methodology is used constructively to support software engineering research. Objective: This study aims to assess the value of mapping studies which are a form of SLR that aims to identify and categorise the available research on a broad software engineering topic. Method: We used a multi-case, participant-observer case study using five examples of studies that were based on preceding mapping studies. We also validated our results by contacting two other researchers who had undertaken studies based on preceding mapping studies and by assessing review comments related to our follow-on studies. Results: Our original case study identified 11 unique benefits that can accrue from basing research on a preceding mapping study of which only two were case specific. We also identified nine problems associated with using preceding mapping studies of which two were case specific. These results were consistent with the information obtained from the validation activities. We did not find an example of an independent research group making use of a mapping study produced by other researchers. Conclusion: Mapping studies can save time and effort for researchers and provide baselines to assist new research efforts. However, they must be of high quality in terms of completeness and rigour if they are to be a reliable basis for follow-on research.

/pdf/using-mapping-studies-as-the-basis-for-further-research-a-30nrr2jq2u.pdf

Using mapping studies as the basis for further research - A participant-observer case study

Survey Errors and Survey Costs

This paper provides a systematic review of previous software fault prediction studies with a specific focus on metrics, methods, and datasets. The review uses 74 software fault prediction papers in 11 journals and several conference proceedings. According to the review results, the usage percentage of public datasets increased significantly and the usage percentage of machine learning algorithms increased slightly since 2005. In addition, method-level metrics are still the most dominant metrics in fault prediction research area and machine learning algorithms are still the most popular methods for fault prediction. Researchers working on software fault prediction area should continue to use public datasets and machine learning algorithms to build better fault predictors. The usage percentage of class-level is beyond acceptable levels and they should be used much more than they are now in order to predict the faults earlier in design phase of software life cycle.

A systematic review of software fault prediction studies

Crop yield prediction using machine learning: A systematic literature review

Software engineering discipline contains several prediction approaches such as test effort prediction, correction cost prediction, fault prediction, reusability prediction, security prediction, effort prediction, and quality prediction. However, most of these prediction approaches are still in preliminary phase and more research should be conducted to reach robust models. Software fault prediction is the most popular research area in these prediction approaches and recently several research centers started new projects on this area. In this study, we investigated 90 software fault prediction papers published between year 1990 and year 2009 and then we categorized these papers according to the publication year. This paper surveys the software engineering literature on software fault prediction and both machine learning based and statistical based approaches are included in this survey. Papers explained in this article reflect the outline of what was published so far, but naturally this is not a complete review of all the papers published so far. This paper will help researchers to investigate the previous studies from metrics, methods, datasets, performance evaluation metrics, and experimental results perspectives in an easy and effective manner. Furthermore, current trends are introduced and discussed.

Cagatay Catal

Papers

A systematic review of software fault prediction studies

Crop yield prediction using machine learning: A systematic literature review

Review: Software fault prediction: A literature review and current trends

Investigating the effect of dataset size, metrics sets, and feature selection techniques on software fault prediction problem

Analysis of transfer learning for deep neural network based plant classification models