Artificial Neural Network approach on Type II Regression Analysis
In this study, the Artificial Neural Network (ANN) approach was applied to the OLS-Bisector technique, which is one of the Type II Regression techniques, through this study. In order to measure the performance of this newly created ANN-Bisector technique, it was compared with the OLS-Bisector techni...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Istanbul University
2021-12-01
|
| Series: | Alphanumeric Journal |
| Subjects: | |
| Online Access: |
http://alphanumericjournal.com/media/Issue/volume-9-issue-2-2021/artificial-neural-network-approach-on-type-ii-regression-analysis.pdf
|
| _version_ | 1797910258782306304 |
|---|---|
| author | Berkalp Tunca Sinan Saraçlı |
| author_facet | Berkalp Tunca Sinan Saraçlı |
| author_sort | Berkalp Tunca |
| collection | DOAJ |
| description | In this study, the Artificial Neural Network (ANN) approach was applied to the OLS-Bisector technique, which is one of the Type II Regression techniques, through this study. In order to measure the performance of this newly created ANN-Bisector technique, it was compared with the OLS-Bisector technique. First of all, literature information on ANN and OLS-Bisector Regression techniques is given, and the features of two techniques are mentioned. In line with this information, a comparison was made between OLS based bisector technique and ANN based bisector techniques. In order to compare these two techniques, they were modeled in different distributions and in different sample sizes. In order to compare the performances of these models, the Mean Absolute Percent Error (MAPE) criterion was used. As a result of the study, it was seen that the ANN based bisector technique gave better results with lower error than the OLS based bisector technique. With this study, it is foreseen that it will represent an example for researchers who want to work in these fields in the future. |
| first_indexed | 2024-04-10T11:21:32Z |
| format | Article |
| id | doaj.art-bb4213962be14890b8bf7fb13b9abaec |
| institution | Directory Open Access Journal |
| issn | 2148-2225 |
| language | English |
| last_indexed | 2024-04-10T11:21:32Z |
| publishDate | 2021-12-01 |
| publisher | Istanbul University |
| record_format | Article |
| series | Alphanumeric Journal |
| spelling | doaj.art-bb4213962be14890b8bf7fb13b9abaec2023-02-15T16:18:36ZengIstanbul UniversityAlphanumeric Journal2148-22252021-12-0192247258http://dx.doi.org/10.17093/alphanumeric.97213821482225Artificial Neural Network approach on Type II Regression AnalysisBerkalp Tunca0Sinan Saraçlı1 None, Afyon Kocatepe University, Department of Statistics, Faculty of Science and Literature In this study, the Artificial Neural Network (ANN) approach was applied to the OLS-Bisector technique, which is one of the Type II Regression techniques, through this study. In order to measure the performance of this newly created ANN-Bisector technique, it was compared with the OLS-Bisector technique. First of all, literature information on ANN and OLS-Bisector Regression techniques is given, and the features of two techniques are mentioned. In line with this information, a comparison was made between OLS based bisector technique and ANN based bisector techniques. In order to compare these two techniques, they were modeled in different distributions and in different sample sizes. In order to compare the performances of these models, the Mean Absolute Percent Error (MAPE) criterion was used. As a result of the study, it was seen that the ANN based bisector technique gave better results with lower error than the OLS based bisector technique. With this study, it is foreseen that it will represent an example for researchers who want to work in these fields in the future. http://alphanumericjournal.com/media/Issue/volume-9-issue-2-2021/artificial-neural-network-approach-on-type-ii-regression-analysis.pdf artificial neural networksmeasurement error modelstype ii regression |
| spellingShingle | Berkalp Tunca Sinan Saraçlı Artificial Neural Network approach on Type II Regression Analysis Alphanumeric Journal artificial neural networks measurement error models type ii regression |
| title | Artificial Neural Network approach on Type II Regression Analysis |
| title_full | Artificial Neural Network approach on Type II Regression Analysis |
| title_fullStr | Artificial Neural Network approach on Type II Regression Analysis |
| title_full_unstemmed | Artificial Neural Network approach on Type II Regression Analysis |
| title_short | Artificial Neural Network approach on Type II Regression Analysis |
| title_sort | artificial neural network approach on type ii regression analysis |
| topic | artificial neural networks measurement error models type ii regression |
| url |
http://alphanumericjournal.com/media/Issue/volume-9-issue-2-2021/artificial-neural-network-approach-on-type-ii-regression-analysis.pdf
|
| work_keys_str_mv | AT berkalptunca artificialneuralnetworkapproachontypeiiregressionanalysis AT sinansaraclı artificialneuralnetworkapproachontypeiiregressionanalysis |