Yazar "Kutluana, Gokhan" seçeneğine göre listele

Listeleniyor 1 - 2 / 2
  • Küçük Resim Yok
    Öğe
    Classification of cardiac disorders using weighted visibility graph features from ECG signals
    (Elsevier Sci Ltd, 2024) Kutluana, Gokhan; Turker, Ilker
    As universal expressions to describe complex systems, graphs are increasingly preferred as a representation method in artificial intelligence. Visibility graphs enable converting time-series data into graph representations, inheriting some key properties of the series. This study investigates the representation capacity of visibility graphs for ECG signals using either the sequence of node weights or the diagonals of the adjacency matrices as feature sets, input to ResNet and Inception classifier models. This approach also reduces the high dimensionality of the original graph representation which features a size of data points squared. Experiments performed on the multi-labeled PTB-XL dataset indicate that the first 3 diagonals of the visibility graph as the feature set to the ResNet model provides superior classification results compared to the original signal, node weights from the visibility graph, or the combinations of these inputs. Having achieved a maximum AUC score of 93.46%, this approach also outperforms the previously recorded ECG classification results for the PTB-XL dataset.
  • Küçük Resim Yok
    Öğe
    TOOL WEAR PREDICTION BY DEEP LEARNING FROM AUGMENTABLE VISIBILITY GRAPH REPRESENTATION OF TIME SERIES DATA
    (Technical Univ Cluj-Napoca, Fac Machine Building Dept Systems Eng, 2023) Turker, Ilker; Tan, Serhat Orkun; Kutluana, Gokhan
    Tool wear prediction has a crucial role for improving manufacturing quality and reliability due to optimizing tool replacement schedules, reducing downtime, and improving overall production efficiency. Deep learning models, having the ability to analyze large and complex datasets, can extract relevant information, and make accurate predictions about the condition of cutting tools. We propose a smart detection methodology based on converting the available sensory data collected from a CNC milling machine into a visibility graph representation. Due to the high dimensionality of the data with 44 attributes related to machining, a multilayer visibility graph representation is achieved after this conversion procedure, resulting in a 44-layered 128x128 adjacency matrix formation. A novel data augmentation technique specifically applicable to graph representation is also employed to increase the data size originally composed of 18 experiments into 360, each one represented as a multilayer graph. Augmented graph representations are further input to a custom CNN deep learning architecture with a split of 70% train, 10% validation and 20% test instances. Results indicate that Augmented Graph-induced classification of CNC mill tool with custom CNN model (GA-CNN) yields full accuracy for detecting whether the tool is worn or not.