Article
Computer Science, Artificial Intelligence
Feng Lin
Summary: The article presents a new learning algorithm that is mathematically equivalent to backpropagation algorithm but eliminates the need for a feedback network, making implementation simpler and increasing biological plausibility for biological neural networks to learn using the new algorithm.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Rongzhao Zhang, Albert C. S. Chung
Summary: This paper introduces a novel CNN quantization framework that compresses deep models to extremely low bitwidth while maintaining high performance. By utilizing an optimized quantizer, radical residual connection scheme, tanh-based derivative function, and distributional loss, the framework achieves superior results compared to state-of-the-art quantization methods, demonstrating lossless performance with ternary quantization on two 3D segmentation datasets.
MEDICAL IMAGE ANALYSIS
(2021)
Article
Engineering, Biomedical
Qiusha Min, Xin Wang, Bo Huang, Zhongwei Zhou
Summary: This paper introduces a medical image data compression technique that combines anatomical information and a custom deep neural network model to improve the efficiency of data transmission and storage. Through dividing specific regions and training optimized predictors, high prediction accuracy and better compression performance than JPEG2000 have been achieved.
BIOMEDICAL SIGNAL PROCESSING AND CONTROL
(2022)
Article
Computer Science, Artificial Intelligence
Jianyong Wang, Lei Zhang, Zhang Yi
Summary: This article proposes a novel mixture convolutional network (MixConvNet) for three-dimensional (3D) medical image segmentation. By utilizing the advantages of 2D convolution and maintaining the learning ability of 3D convolution, MixConvNet achieves a trade-off between efficiency and accuracy. Experimental results show that MixConvNet outperforms other state-of-the-art methods.
INTERNATIONAL JOURNAL OF NEURAL SYSTEMS
(2023)
Article
Computer Science, Information Systems
Yetian Fan, Wenyu Yang
Summary: This paper proposes a BP algorithm with graph regularization (BPGR) to optimize the parameters and improve the generalization performance of BP neural networks. The proposed method enforces the latent features of the hidden layer to be more concentrated, enhancing the network's generalization capability. The modified graph regularization simplifies gradient calculation and better penalizes extreme weight values. Additionally, the graph regularization can be integrated with deep neural networks to further improve their generalization performance.
INFORMATION SCIENCES
(2022)
Article
Computer Science, Artificial Intelligence
Sathishkumar Karupusamy, J. Refonaa, Sakthidasan Sankaran, Priyanka Dahiya, Mohd Anul Haq, Anil Kumar
Summary: The development of IoT systems has created higher demands for processing and storage environment, as well as raised issues such as energy consumption and data compression. This article proposes an approach that utilizes data mining to optimize energy usage and compress data, validated through the analysis of driving behavior.
Article
Multidisciplinary Sciences
Tariq M. Khan, Syed S. Naqvi, Erik Meijering
Summary: Recent advancements in encoder-decoder neural network architecture design have improved the performance of medical image segmentation tasks. However, these state-of-the-art networks may be too computationally demanding for affordable hardware, leading to the need for practical modifications. This study investigates the effects of downsampling input images and reducing network depth or size on segmentation performance, finding that image complexity can guide the selection of the best approach for a given dataset. Experimental results suggest that median frequency is the most suitable complexity measure when choosing an acceptable input downsampling factor and determining the use of a deep versus shallow, large-size versus lightweight network. For high-complexity datasets, a lightweight network running on the original images may yield better results, while the opposite may be true for low-complexity images.
SCIENTIFIC REPORTS
(2022)
Article
Chemistry, Multidisciplinary
Jie Zhao, Maolin Zhang, Chenchen Wang, Zheng Mao, Yizhong Zhang
Summary: The pore types of intersalt shale reservoirs are diverse and complex, with sizes ranging from nanometers to microns. Image segmentation using digital core technology is crucial for studying the micropore structures. By adopting the backpropagation neural network segmentation method with a genetic algorithm, the study shows improved accuracy and efficiency in pore prediction, making it closer to gas-measured values.
Article
Computer Science, Information Systems
Guoqing Li, Meng Zhang, Jiuyang Wang, Dongpeng Weng, Henk Corporaal
Summary: This paper proposes a simple shared channel weight convolution (SCWC) approach to reduce the number of parameters in convolutional neural networks (CNNs). Experimental results demonstrate the effectiveness of SCWC in image classification and object detection tasks.
INFORMATION SCIENCES
(2022)
Article
Computer Science, Artificial Intelligence
Haifeng Wang, Daehan Won, Sang Won Yoon
Summary: The study introduces an adaptive neural architecture optimization model that optimizes convolutional neural network structure using integer programming to improve model accuracy and convergence speed. The model considers accuracy and training trends, evaluates candidate models' performance using recurrent neural network, and utilizes a heuristic process for optimization.
APPLIED SOFT COMPUTING
(2021)
Article
Computer Science, Artificial Intelligence
Chilukamari Rajesh, Sushil Kumar
Summary: The study proposed an automatic network evolution model based on Differential Evolution for optimizing medical image denoising network structures and hyperparameters. By exploring the fittest parameters and accelerating the training process, the model was evaluated on four different medical image datasets.
APPLIED SOFT COMPUTING
(2022)
Article
Computer Science, Artificial Intelligence
Baichen Liu, Zhi Han, Xi'ai Chen, Wenming Shao, Huidi Jia, Yanmei Wang, Yandong Tang
Summary: Convolutional neural networks (CNNs) can be inconvenient in situations with limited storage space due to their numerous parameters. This paper proposes a novel compact design for convolutional layers using spatial transformation to achieve a lower-rank form. The effectiveness of the method is validated in an image classification task.
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Computer Science, Software Engineering
Yi Xiao, Jin Wu, Jie Zhang, Peiyao Zhou, Yan Zheng, Chi-Sing Leung, Ladislav Kavan
Summary: This article introduces a deep learning-based method for generating colored images, which allows users to control the results through global and local inputs. The authors propose a two-stage deep colorization method and design a loss function to differentiate the influences of different inputs. They also propose the application of color theme recommendation and image compression scheme.
IEEE TRANSACTIONS ON VISUALIZATION AND COMPUTER GRAPHICS
(2022)
Article
Plant Sciences
Kanae Masuda, Eriko Kuwada, Maria Suzuki, Tetsuya Suzuki, Takeshi Niikawa, Seiichi Uchida, Takashi Akagi
Summary: Deep neural network techniques are widely used in plant biology, particularly in phenotyping studies. This study combines explainable CNN models and transcriptomic analysis to provide a physiological interpretation of rapid over-softening in persimmon fruit, uncovering a novel aspect of fruit premonitory reactions.
PLANT AND CELL PHYSIOLOGY
(2023)
Article
Computer Science, Artificial Intelligence
Sahadev Poudel, Sang-Woong Lee
Summary: Automatic segmentation of medical images is a challenging task due to the diverse characteristics of polyps or tumors. This study addresses the limitations of existing deep learning models by capturing multi-scale global features and utilizing an attention mechanism to improve segmentation accuracy and suppress noise. The proposed method outperforms baseline models across different datasets and backbone architectures, enhancing segmentation quality and overall performance.
APPLIED SOFT COMPUTING
(2021)
Article
Computer Science, Artificial Intelligence
Adil Amirjanov, Kamil Dimililer
IET IMAGE PROCESSING
(2019)
Article
Computer Science, Hardware & Architecture
Kamil Dimililer, Hilmi Dindar, Fadi Al-Turjman
Summary: Earthquakes in the Eastern Mediterranean are mostly tectonic, with depths generally varying between 0 and 30 km. As technology has advanced, the use of IoT and machine learning have proven effective in predicting and preparing for future natural disasters.
MICROPROCESSORS AND MICROSYSTEMS
(2021)
Article
Computer Science, Hardware & Architecture
Hanifa Teimourian, Amir Teimourian, Kamil Dimililer, Fadi Al-Turjman
Summary: In this paper, a framework for wind energy harvesting based on the Internet of Things (IoT) was investigated in the historical city of Bam. Results showed that peak power density and energy density occur during the summer season, making it advantageous for Bam city with higher energy demand in the hot summer months. Additionally, an economic assessment was conducted to determine the feasibility of installing small-scale wind turbines in the city.
JOURNAL OF SUPERCOMPUTING
(2022)
Article
Engineering, Electrical & Electronic
Kamil Dimililer
Summary: Machine learning algorithms are trained to relate medical image contents to compression ratios, with the radial basis function neural network achieving the best classification results for optimal compression ratios for X-ray images.
SIGNAL IMAGE AND VIDEO PROCESSING
(2022)
Article
Multidisciplinary Sciences
Kamil Dimililer, Boran Sekeroglu
Summary: In the past decade, the use of computer-aided diagnosis (CAD) and deep learning models for skin lesion analysis has become common. This study presents a transfer learning model using Convolutional Neural Networks (CNN) to classify skin lesion images obtained by smartphones. The designed model increased the classification rates by 20% compared to conventional CNN. The transfer learning model achieved high recall, specificity, and accuracy in detecting cancerous lesions, and high recall, precision, and F1 score in classifying six skin lesions, demonstrating the efficacy of transfer learning in skin lesion diagnosis.
GAZI UNIVERSITY JOURNAL OF SCIENCE
(2023)
Article
Computer Science, Artificial Intelligence
Kamil Dimililer, Hanifa Teimourian, Fadi Al-Turjman
Summary: This paper discusses the use of machine learning algorithms in classifying and comparing high-redshift radio galaxies. The results show that different machine learning algorithms have different accuracy rates when applied to different datasets.
JOURNAL OF EXPERIMENTAL & THEORETICAL ARTIFICIAL INTELLIGENCE
(2022)
Article
Computer Science, Hardware & Architecture
Devrim Kayali, Nemah Abu Shama, Suleyman Asir, Kamil Dimililer
Summary: Iron, as a trace element, has a crucial role in the human immune system, particularly in combating SARS-CoV-2 variants. Electrochemical methods, such as square wave voltammetry (SQWV) and differential pulse voltammetry (DPV), are convenient for detecting various compounds, including heavy metals, due to their simplicity. This study improved machine learning models to classify concentrations of an analyte based solely on obtained voltammograms. SQWV and DPV were used to quantify the concentrations of ferrous ions (Fe+2) in potassium ferrocyanide (K4Fe(CN)(6)), and the machine learning models validated the data classifications. The highest accuracy of 100% was achieved for each analyte in 25 seconds using our models, outperforming previously used algorithms for data classification.
JOURNAL OF SUPERCOMPUTING
(2023)
Article
Computer Science, Information Systems
Boran Sekeroglu, Yoney Kirsal Ever, Kamil Dimililer, Fadi Al-Turjman
Summary: This study aims to analyze the performance of machine learning models on different datasets, considering various training strategies and evaluation metrics. The results demonstrate that the deep Long-Short Term Memory (LSTM) neural network outperforms other models and indicate the significant impact of cross-validation on the experimental results.
Review
Computer Science, Artificial Intelligence
B. Sekeroglu, K. Dimililer
NEURAL NETWORK WORLD
(2020)
Proceedings Paper
Computer Science, Artificial Intelligence
Kamil Dimililer, Yoney Kirsal Ever, Sipan Masoud Mustafa
10TH INTERNATIONAL CONFERENCE ON THEORY AND APPLICATION OF SOFT COMPUTING, COMPUTING WITH WORDS AND PERCEPTIONS - ICSCCW-2019
(2020)
Article
Education & Educational Research
Boran Sekeroglu, Kamil Dimililer, Kubra Tuncal
DILEMAS CONTEMPORANEOS-EDUCACION POLITICA Y VALORES
(2019)
Proceedings Paper
Computer Science, Theory & Methods
Boran Sekeroglu, Kamil Dimililer, Kubra Tuncal
PROCEEDINGS OF 2019 8TH INTERNATIONAL CONFERENCE ON EDUCATIONAL AND INFORMATION TECHNOLOGY (ICEIT 2019)
(2019)
Article
Education & Educational Research
Kamil Dimililer, Getinet Amare Mekonnen
DILEMAS CONTEMPORANEOS-EDUCACION POLITICA Y VALORES
(2019)
Proceedings Paper
Computer Science, Artificial Intelligence
Kamil Dimililer, Idoko John Bush
MAN-MACHINE INTERACTIONS 5, ICMMI 2017
(2018)
Proceedings Paper
Computer Science, Interdisciplinary Applications
Ezekiel T. Ogidan, Kamil Dimililer, Yoney Kirsal Ever
2018 2ND INTERNATIONAL SYMPOSIUM ON MULTIDISCIPLINARY STUDIES AND INNOVATIVE TECHNOLOGIES (ISMSIT)
(2018)