On learning algorithm selection for classification

Ali, Shawkat; Smith, Kate

doi:10.1016/j.asoc.2004.12.002

Cited by 317 publications

(189 citation statements)

References 11 publications

Supporting

Mentioning

187

Contrasting

Unclassified

Order By: Relevance

“…Here, three types of training algorithms having eight training functions have been evaluated for classification of brain hematoma. They are Gradient Descent algorithms (traingd, traingdm, trainrp), Conjugate Gradient algorithms (trainscg, traincgf, traincgp), Quasi-Newton algorithms (trainbfg,trainlm) [8].…”

Section: Training Algorithmsmentioning

confidence: 99%

Comparison of Neural Network Training Functions for Hematoma Classification in Brain CT Images

Sharma¹,

Venugopalan²

2014

IOSRJCE

106

View full text Add to dashboard Cite

show abstract

Section: Training Algorithmsmentioning

confidence: 99%

Comparison of Neural Network Training Functions for Hematoma Classification in Brain CT Images

Sharma¹,

Venugopalan²

2014

IOSRJCE

106

View full text Add to dashboard Cite

show abstract

“…Hence, performance measurement is essential for model selection, i.e. to identify the most suited classification technique as well as to tune the respective parameters (Ali and Smith, 2006). It has been shown that traditional performance measures such as the Gini coefficient, the KS statistic, and the AUC measure are inappropriate in many cases and may lead to incorrect conclusions (Hand, 2005(Hand, , 2009), since they do not always properly take into account the business reality of credit scoring.…”

Section: Introductionmentioning

confidence: 99%

Development and application of consumer credit scoring models using profit-based classification measures

Verbraken¹,

Bravo²,

Weber³

et al. 2014

European Journal of Operational Research

151

110

View full text Add to dashboard Cite

This paper presents a new approach for consumer credit scoring, by tailoring a profit-based classification performance measure to credit risk modeling. This performance measure takes into account the expected profits and losses of credit granting and thereby better aligns the model developers' objectives with those of the lending company. It is based on the Expected Maximum Profit (EMP) measure and is used to find a trade-off between the expected losses -driven by the exposure of the loan and the loss given default -and the operational income given by the loan. Additionally, one of the major advantages of * NOTICE: this is the author's version of a work that was accepted for publication in the European Journal of Operational Research. Changes resulting from the publishing process, such as peer review, editing, corrections, structural formatting, and other quality control mechanisms may not be reflected in this document. Changes may have been made to this work since it was submitted for publication. Please cite this paper as follows: Verbraken, T., Bravo, C., Weber, R. and Baesens, B. (2014) Development and application of consumer credit scoring models using profit-based classification measures. European Journal of Operational Research. In Press. Available Online: http://www.sciencedirect.com/ science/article/pii/S0377221714003105 † Email Addresses: thomas.verbraken@kuleuven.be, crbravo@utalca.cl (corresponding), rweber@dii.uchile.cl, bart.baesens@kuleuven.be 1 using the proposed measure is that it permits to calculate the optimal cutoff value, which is necessary for model implementation. To test the proposed approach, we use a dataset of loans granted by a government institution, and benchmarked the accuracy and monetary gain of using EMP, accuracy, and the area under the ROC curve as measures for selecting model parameters, and for determining the respective cutoff values. The results show that our proposed profit-based classification measure outperforms the alternative approaches in terms of both accuracy and monetary value in the test set, and that it facilitates model deployment.

show abstract

“…MLAs from five classes are chosen so as to introduce a certain degree of variation among the classifiers, i.e. they should not make identical or correlated errors [13]. The justification behind choosing these MLAs are next discussed.…”

Section: Machine Learningmentioning

confidence: 99%

Early Prediction of Students Performance using Machine Learning Techniques

Acharya¹,

Sinha²

2014

IJCA

View full text Add to dashboard Cite

In recent years Educational Data Mining (EDM) has emerged as a new field of research due to the development of several statistical approaches to explore data in educational context. One such application of EDM is early prediction of student results. This is necessary in higher education for identifying the "weak" students so that some form of remediation may be organized for them. In this paper a set of attributes are first defined for a group of students majoring in Computer Science in some undergraduate colleges in Kolkata. Since the numbers of attributes are reasonably high, feature selection algorithms are applied on the data set to reduce the number of features. Five classes of Machine Learning Algorithm (MLA) are then applied on this data set and it was found that the best results were obtained with the decision tree class of algorithms. It was also found that the prediction results obtained with this model are comparable with other previously developed models.

show abstract

On learning algorithm selection for classification

Cited by 317 publications

References 11 publications

Comparison of Neural Network Training Functions for Hematoma Classification in Brain CT Images

Comparison of Neural Network Training Functions for Hematoma Classification in Brain CT Images

Development and application of consumer credit scoring models using profit-based classification measures

Early Prediction of Students Performance using Machine Learning Techniques

Contact Info

Product

Resources

About