Software Effort Estimation using Machine Learning Techniques with Robust Confidence Intervals

Braga, Petrônio L.; Oliveira, Adriano L. I.; Meira, Sílvio R. L.

doi:10.1109/his.2007.56

Cited by 24 publications

(6 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…There are additional theoretical and practical implications that can be drawn from this research. This new framework can be applied to predict risk probability in the same context as the prediction of activity duration in projects as was suggested in (Braga et al 2008), using neural networks tools. Their study concluded that using integrated machine learning techniques improved the reliability and accuracy of project duration forecasts.…”

Section: Discussionmentioning

confidence: 99%

“…For example, using analytical tools in big data systems to detect fraud by rogue suppliers to reduce risks in the supply chain and increase its reliability (Zage et al 2013). The ability of machine learning techniques and methods to predict the duration of activities in projects was demonstrated in (Braga et al 2008), using several methods including decision tree, support vector machines, bagging-bootstrap aggregating predictors, and neural networks. The study concludes that using integrated machine learning techniques improved the reliability and accuracy of project duration forecasts.…”

Section: New Tools and Frameworkmentioning

confidence: 99%

See 1 more Smart Citation

Deconstructing Risk Factors for Predicting Risk Assessment in Supply Chains Using Machine Learning

Burstein

Zuckerman

2023

JRFM

View full text Add to dashboard Cite

Risk management is an ongoing process that includes several stages of mapping and identification, analysis, and evaluation, planning, and implementation to reduce risks and ensure ongoing control. Risk management along the supply chains has become more significant in recent years due to an increased complexity of the relationships between components in the chain as well as various disruptions such as climate change, COVID-19, or geo-political scenarios. The current literature alongside the increase in complexity and frequency of risk events, leads us to the single, most prominent challenge in risk management today: the auditor’s subjectivity in determining the risk levels. Simply stated, two different auditors may assess a given situation differently due to their specific history and experience. Specifically, it seems to be extremely difficult to find cases in which different auditors, working on the same organization, made the same risk assessment. With that in mind, this research aims to reduce the human subjectivity bias and reach a risk evaluation that is as objective as possible, by using the machine learning approach. For this aim the paper introduces a new risk assessment framework based on factors analysis and artificial neural network as the predictive model. We first introduced a new approach of deconstructing the risk factors into their basic elements and analyzing them as a feature vector. Next, we collected unique, real-world data of risk surveys and audit reports from 60 industrial companies of various industries (from plastic and metal factories to logistic and medical devices companies). Lastly, we constructed a neural network to predict the risk levels of operational processes in the industry. We trained our model on 42 samples and managed to achieve a R2 score of 0.9 on the test set of 18 samples. Our model was validated and managed to predict the risk accuracy with R = 0.95 in accordance with the human auditor results.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: New Tools and Frameworkmentioning

confidence: 99%

Deconstructing Risk Factors for Predicting Risk Assessment in Supply Chains Using Machine Learning

Burstein

Zuckerman

2023

JRFM

View full text Add to dashboard Cite

show abstract

“…The original dataset consists of 12 attributes but in this study the ProjectID attribute was omitted from the original dataset because it has no meaning to the study, the left are ten independent attributes and one dependent attribute (effort), all the values in this dataset are numeric but only one nominal attribute that is Language. Many researchers used this dataset in their experiments including [1], [8], [10] and many others. Despite the fact that this dataset is now more than 25 years old, it is one of the largest and most used publicly available datasets [8].…”

Section: Methodsmentioning

confidence: 99%

Software Development Effort Estimation Using Ensemble Machine Learning

2017

IJCCIE

View full text Add to dashboard Cite

Abstract-In software engineering, the main aim is to develop a high quality project that fall within scheduled time and budget, this procedure is called effort estimation. Effort estimation is crucial and important for a company to do because hiring more people than needed will lead to loss of income, and hiring less people than needed will lead to delay of project delivery. The aim of this study is to estimate software effort objectively by using machine learning techniques instead of subjective and time consuming estimation methods. Models using two machine learning techniques which are Support Vector Machine (SVM) and K-Nearest Neighbor (k-NN) separately and combining those together using ensemble learning were tried on two public datasets namely Desharnais and Maxwell. Results show that svm technique outperform k-nn technique, also ensemble learning improves the results.

show abstract

“…Confidence intervals seem the most suitable measure for estimate variation in SDEE since they are specific to error context as they assess the reliability of a probability distribution independently of its shape . For example, Stamelos et al used confidence intervals as a performance measure for their estimate variation approach using bootstrapping over a portfolio of projects, while Braga et al used confidence intervals for describing estimate variation using machine learning estimation techniques. Based on the Gaussian distribution properties shown in Figure , we defined three predicted intervals that mimic the relevant Gaussian intervals:

I 1 = ([],, italicEest - ∆ italiceff, italicEest + ∆ italiceff)

I 2 = ([],, italicEest - 2 \times ∆ italiceff, italicEest + 2 \times ∆ italiceff)

I 3 = ([],, italicEest - 3 \times ∆ italiceff, italicEest + 3 \times ∆ italiceff)

where Eest is the project estimated effort and ∆eff is the estimated deviation.…”

Section: Performance Measures Datasets Description and Effort Estimmentioning

confidence: 99%

Empirical evaluation of an entropy‐based approach to estimation variation of software development effort

Koutbi

Idri

Abran

2018

J Software Evolu Process

View full text Add to dashboard Cite

As effort estimation has gained increasing attention, most of the techniques proposed have focused on the accuracy of effort estimates. Yet no clear conclusions have been drawn on which techniques perform best in all contexts. We propose an entropy‐based approach to effort estimate variation caused by measurement and model error sources whatever the effort estimation technique used. The proposed approach was empirically evaluated by exploring three entropy formulae, four interpolation methods, and two analogy‐based effort estimation approaches (crisp and fuzzy analogy) over seven datasets using the Jackknife evaluation method. The obtained results show that the three entropy formulae have in general the same positive influence on the performance of the entropy‐based approach measured in terms of absolute error of effort deviation. In addition, the spline interpolation outperformed all other interpolation methods, using any of the entropy formulae. Moreover, achievement percentages of the best variants of our approach closely approximated those of the Gaussian distribution confirming that the Gaussian distribution is useful for characterizing effort estimate variation.

show abstract

Software Effort Estimation using Machine Learning Techniques with Robust Confidence Intervals

Cited by 24 publications

References 12 publications

Deconstructing Risk Factors for Predicting Risk Assessment in Supply Chains Using Machine Learning

Deconstructing Risk Factors for Predicting Risk Assessment in Supply Chains Using Machine Learning

Software Development Effort Estimation Using Ensemble Machine Learning

Empirical evaluation of an entropy‐based approach to estimation variation of software development effort

Contact Info

Product

Resources

About