Matches in SemOpenAlex for { <https://semopenalex.org/work/W4316021475> ?p ?o ?g. }
Showing items 1 to 79 of
79
with 100 items per page.
- W4316021475 abstract "For financial services to be competitive, artificial intelligence and machine learning have become essential. This study benchmarked Deep Neural Networks (DNN) and Gradient Boosting Machines (GBM) based on their superior model accuracy performance in most previous studies. Consequently, this study aims to perform a comparative analysis of classification power between DNN and GBM. To date, there is no such comparison. Each model underwent extensive tuning to achieve performance appropriate for the comparison in this study. The author deployed three renowned credit risk datasets for model construction to train GBM, DNN + ReLU, DNN + Maxout, and DNN + Tanh classifiers. For the model evaluation, the area under the receiver operating characteristics curve (AUC ROC) was utilized to assess the models' accuracy performance. The assessment also includes investigating the effect of splitting the training/test sets using the split method of 80:20 split and 70:30 split. The results indicate that GBM tends to be faster and more potent than DNN due to its lower processing requirements. GBM is hence the preferred method for credit scoring prediction. Overall, this study observed that DNN and GBM could be the cutting-edge for solving classification problems. Nevertheless, GBM should be the first-choice option due to its simplicity, quicker to train, and higher accuracy." @default.
- W4316021475 created "2023-01-14" @default.
- W4316021475 creator A5042623204 @default.
- W4316021475 date "2022-12-04" @default.
- W4316021475 modified "2023-10-18" @default.
- W4316021475 title "Assessment of Deep Neural Network and Gradient Boosting Machines for Credit Risk Prediction Accuracy" @default.
- W4316021475 cites W1678356000 @default.
- W4316021475 cites W1901616594 @default.
- W4316021475 cites W1980264541 @default.
- W4316021475 cites W1988790447 @default.
- W4316021475 cites W2023003828 @default.
- W4316021475 cites W2085988980 @default.
- W4316021475 cites W2131816657 @default.
- W4316021475 cites W2257979135 @default.
- W4316021475 cites W2470019544 @default.
- W4316021475 cites W2605486874 @default.
- W4316021475 cites W2766447205 @default.
- W4316021475 cites W2787468747 @default.
- W4316021475 cites W2788025656 @default.
- W4316021475 cites W2892515961 @default.
- W4316021475 cites W2904485001 @default.
- W4316021475 cites W2919115771 @default.
- W4316021475 cites W2929399345 @default.
- W4316021475 cites W2934302500 @default.
- W4316021475 cites W2949829362 @default.
- W4316021475 cites W2951781852 @default.
- W4316021475 cites W2969625533 @default.
- W4316021475 cites W2975867066 @default.
- W4316021475 cites W2977203305 @default.
- W4316021475 cites W3085480867 @default.
- W4316021475 cites W3135286407 @default.
- W4316021475 cites W3148119887 @default.
- W4316021475 cites W3176827378 @default.
- W4316021475 cites W3180689527 @default.
- W4316021475 cites W3191198889 @default.
- W4316021475 cites W4210754757 @default.
- W4316021475 cites W4212883601 @default.
- W4316021475 cites W4244945522 @default.
- W4316021475 cites W4283259653 @default.
- W4316021475 doi "https://doi.org/10.1109/cicn56167.2022.10008264" @default.
- W4316021475 hasPublicationYear "2022" @default.
- W4316021475 type Work @default.
- W4316021475 citedByCount "0" @default.
- W4316021475 crossrefType "proceedings-article" @default.
- W4316021475 hasAuthorship W4316021475A5042623204 @default.
- W4316021475 hasConcept C108583219 @default.
- W4316021475 hasConcept C119857082 @default.
- W4316021475 hasConcept C154945302 @default.
- W4316021475 hasConcept C169258074 @default.
- W4316021475 hasConcept C41008148 @default.
- W4316021475 hasConcept C46686674 @default.
- W4316021475 hasConcept C50644808 @default.
- W4316021475 hasConcept C58471807 @default.
- W4316021475 hasConcept C70153297 @default.
- W4316021475 hasConceptScore W4316021475C108583219 @default.
- W4316021475 hasConceptScore W4316021475C119857082 @default.
- W4316021475 hasConceptScore W4316021475C154945302 @default.
- W4316021475 hasConceptScore W4316021475C169258074 @default.
- W4316021475 hasConceptScore W4316021475C41008148 @default.
- W4316021475 hasConceptScore W4316021475C46686674 @default.
- W4316021475 hasConceptScore W4316021475C50644808 @default.
- W4316021475 hasConceptScore W4316021475C58471807 @default.
- W4316021475 hasConceptScore W4316021475C70153297 @default.
- W4316021475 hasLocation W43160214751 @default.
- W4316021475 hasOpenAccess W4316021475 @default.
- W4316021475 hasPrimaryLocation W43160214751 @default.
- W4316021475 hasRelatedWork W3130659594 @default.
- W4316021475 hasRelatedWork W3151529617 @default.
- W4316021475 hasRelatedWork W3159988495 @default.
- W4316021475 hasRelatedWork W3200719183 @default.
- W4316021475 hasRelatedWork W4220785415 @default.
- W4316021475 hasRelatedWork W4281616679 @default.
- W4316021475 hasRelatedWork W4288057626 @default.
- W4316021475 hasRelatedWork W4292969247 @default.
- W4316021475 hasRelatedWork W4293069612 @default.
- W4316021475 hasRelatedWork W4313488044 @default.
- W4316021475 isParatext "false" @default.
- W4316021475 isRetracted "false" @default.
- W4316021475 workType "article" @default.