Keleşoğlu, Nur

Profile Picture
Profile URL
https://gcris.yasar.edu.tr/handle/123456789/12754
Name Variants
Job Title
Araş.Gör.
Email Address
Main Affiliation
01.01.09.02. Elektrik- Elektronik Mühendisliği
Status
Former Staff
Website
ORCID ID
0000-0002-0306-7281
Scopus Author ID
Turkish CoHE Profile ID
Google Scholar ID
WoS Researcher ID
HSD-6958-2023

Sustainable Development Goals

SDG data is not available
This researcher does not have a Scopus ID.
Documents

4

Citations

9

Go to WoS profile
Scholarly Output

1

Articles

0

Views / Downloads

0/0

Supervised MSc Theses

1

Supervised PhD Theses

0

WoS Citation Count

0

Scopus Citation Count

0

Patents

0

Projects

0

WoS Citations per Publication

0.00

Scopus Citations per Publication

0.00

Open Access Source

0

Supervised Theses

1

Journals data is not available

Scopus Quartile Distribution

Quartile distribution chart data is not available

Competency Cloud

GCRIS Competency Cloud

Filters

20231
Reset filters

Settings

Scholarly Output Search Results

Now showing 1 - 1 of 1
  • Thumbnail Image
    Master Thesis
    Destek vektör regresyonu için örnek tabanlı düzenleme
    (2023) Keleşoğlu, Nur; Güzeliş, Cüneyt
    Regression analysis is a statistical method used in machine learning to estimate the relationship between two or more quantitative variables in business, finance, economics, engineering, and other disciplines. In this thesis, we proposed a sample based regularization algorithm that augments the cost function of Support Vector Regression (SVR) to increase generalization ability. We aim to increase the generalization capability of SVR, which is a Support Vector Machine (SVM) based machine learning model generally used for regression problems. Contrary to the current approaches to the generalization ability problem, we have improved the performance of the SVR model by increasing the generalization ability of the model. We reduce the weights towards zero, decrease the number of support vectors of the SVR model and increase the generalization ability of the model by the proposed method. In order to see the effect of our approach on the performance of the model, we compared it with the conventional regression machine learning models. We also compared the results of the proposed model with the results of a recent study. In order to evaluate the performance of our approach, we compared the R2 score metric, Mean Square Error (MSE), Root Mean Squared Error (RMSE), and Mean Absolute Error (MAE) error metrics. We also present the training and execution times of the models. Moreover, we show the implementation of the SVR model with a sample based regularization in indoor fire and forest fire detection applications. Accordingly, we develop a Hybrid SR-SVR architecture with the novel sample based regularization for SVR. We compared this method with the machine learning models. In addition, we present our results in terms of Accuracy, Precision, Recall, F1 score, True Positive Rate (TPR), and True Negative Rate (TNR). The study presented in this thesis contributes to the Support Vector Regression literature by introducing an efficient regularization method, so called sample based regularization.