A novel performance metric for building an optimized classifier
Problem statement: Typically, the accuracy metric is often applied for optimizing the heuristic or stochastic classification models. However, the use of accuracy metric might lead the searching process to the sub-optimal solutions due to its less discriminating values and it is also not robust to th...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Science Publications
2011
|
| Online Access: | http://psasir.upm.edu.my/id/eprint/22462/ http://psasir.upm.edu.my/id/eprint/22462/1/jcssp.2011.582.590.pdf |
| _version_ | 1848844491230806016 |
|---|---|
| author | Hossin, Mohammad Sulaiman, Md. Nasir Mustapha, Aida Mustapha, Norwati |
| author_facet | Hossin, Mohammad Sulaiman, Md. Nasir Mustapha, Aida Mustapha, Norwati |
| author_sort | Hossin, Mohammad |
| building | UPM Institutional Repository |
| collection | Online Access |
| description | Problem statement: Typically, the accuracy metric is often applied for optimizing the heuristic or stochastic classification models. However, the use of accuracy metric might lead the searching process to the sub-optimal solutions due to its less discriminating values and it is also not robust to the changes of class distribution. Approach: To solve these detrimental effects, we propose a novel performance metric which combines the beneficial properties of accuracy metric with the extended recall and precision metrics. We call this new performance metric as Optimized Accuracy with Recall-Precision (OARP). Results: In this study, we demonstrate that the OARP metric is theoretically better than the accuracy metric using four generated examples. We also demonstrate empirically that a naïve stochastic classification algorithm, which is Monte Carlo Sampling (MCS) algorithm trained with the OARP metric, is able to obtain better predictive results than the one trained with the conventional accuracy metric. Additionally, the t-test analysis also shows a clear advantage of the MCS model trained with the OARP metric over the accuracy metric alone for all binary data sets. Conclusion: The experiments have proved that the OARP metric leads stochastic classifiers such as the MCS towards a better training model, which in turn will improve the predictive results of any heuristic or stochastic classification models. |
| first_indexed | 2025-11-15T08:31:46Z |
| format | Article |
| id | upm-22462 |
| institution | Universiti Putra Malaysia |
| institution_category | Local University |
| language | English |
| last_indexed | 2025-11-15T08:31:46Z |
| publishDate | 2011 |
| publisher | Science Publications |
| recordtype | eprints |
| repository_type | Digital Repository |
| spelling | upm-224622016-06-08T08:58:37Z http://psasir.upm.edu.my/id/eprint/22462/ A novel performance metric for building an optimized classifier Hossin, Mohammad Sulaiman, Md. Nasir Mustapha, Aida Mustapha, Norwati Problem statement: Typically, the accuracy metric is often applied for optimizing the heuristic or stochastic classification models. However, the use of accuracy metric might lead the searching process to the sub-optimal solutions due to its less discriminating values and it is also not robust to the changes of class distribution. Approach: To solve these detrimental effects, we propose a novel performance metric which combines the beneficial properties of accuracy metric with the extended recall and precision metrics. We call this new performance metric as Optimized Accuracy with Recall-Precision (OARP). Results: In this study, we demonstrate that the OARP metric is theoretically better than the accuracy metric using four generated examples. We also demonstrate empirically that a naïve stochastic classification algorithm, which is Monte Carlo Sampling (MCS) algorithm trained with the OARP metric, is able to obtain better predictive results than the one trained with the conventional accuracy metric. Additionally, the t-test analysis also shows a clear advantage of the MCS model trained with the OARP metric over the accuracy metric alone for all binary data sets. Conclusion: The experiments have proved that the OARP metric leads stochastic classifiers such as the MCS towards a better training model, which in turn will improve the predictive results of any heuristic or stochastic classification models. Science Publications 2011 Article PeerReviewed application/pdf en http://psasir.upm.edu.my/id/eprint/22462/1/jcssp.2011.582.590.pdf Hossin, Mohammad and Sulaiman, Md. Nasir and Mustapha, Aida and Mustapha, Norwati (2011) A novel performance metric for building an optimized classifier. Journal of Computer Science, 7 (4). pp. 582-590. ISSN 1549-3636; ESSN: 1552-6607 http://www.thescipub.com/abstract/10.3844/jcssp.2011.582.590 10.3844/jcssp.2011.582.590 |
| spellingShingle | Hossin, Mohammad Sulaiman, Md. Nasir Mustapha, Aida Mustapha, Norwati A novel performance metric for building an optimized classifier |
| title | A novel performance metric for building an optimized classifier |
| title_full | A novel performance metric for building an optimized classifier |
| title_fullStr | A novel performance metric for building an optimized classifier |
| title_full_unstemmed | A novel performance metric for building an optimized classifier |
| title_short | A novel performance metric for building an optimized classifier |
| title_sort | novel performance metric for building an optimized classifier |
| url | http://psasir.upm.edu.my/id/eprint/22462/ http://psasir.upm.edu.my/id/eprint/22462/ http://psasir.upm.edu.my/id/eprint/22462/ http://psasir.upm.edu.my/id/eprint/22462/1/jcssp.2011.582.590.pdf |