Addressing overfitting in comparative study for deep learning-based classification

Addressing overfitting in comparative study for deep learning-based classification

Deep learning
Feature extraction
Few-shot learning
InceptionV3
Overfitting
Stanford dog
Xception

Despite significant advancements in deep learning methodologies for animal species classification, there remains a notable research gap in effectively addressing biases inherent in training datasets, combating overfitting during model training, and enhancing overall performance to ensure reliable and accurate classification results in real-world applications. Therefore, this study explores the complex challenges of dog species classification, with a specific focus on addressing biases, combatting overfitting, and enhancing overall performance using deep learning methodologies. Initially, the Stanford Dog dataset serves as the foundation for training, complemented by additional data from annotated datasets. The primary aim is to mitigate biases and reduce overfitting, which is essential for improving the performance of deep learning-based classification in terms of dataset size and computational time. Feature extraction and few-shot learning techniques are compared to assess and improve the model performance. The experimentation involves the utilization of optimal classifiers, specifically InceptionV3 and Xception. In order to tackle overfitting, a range of strategies are deployed, including data augmentation, early stopping, and the integration of dropout and freezing layers which particularly achieved a better performance with Xception on the augmented dataset.

Jing-Yee Ong, Lee-Yeng Ong, Meng-Chew Leow

Journal homepage: http://journal.uad.ac.id/index.php/TELKOMNIKA

Mar 11, 2025

PERI IRAWAN

PDF

ENGLISH

TEXT