Indonesian Journal of Electrical Engineering and Computer Science
Not a member yet
9109 research outputs found
Sort by
Improved deep learning architecture for skin cancer classification
Get PDFA leading cause of mortality globally, skin cancer is deadly. Early skin cancer diagnosis reduces mortality. Visual inspection is the main skin cancer diagnosis tool; however, it is imprecise. Researchers propose deep-learning techniques to assist physicians identify skin tumors fast and correctly. Deep convolutional neural networks (CNNs) can identify distinct objects in complex tasks. We train a CNN on photos with merely pixels and illness labels to classify skin lesions. We train on HAM-10000 using a CNN. On the HAM10000 dataset, the suggested model scored 95.23% efficiency, 95.30% sensitivity, and 95.91% specificity
Software component selection methods and techniques: a systematic review
Get PDFSoftware component selection is critical in software engineering due to its vital role in reducing software development cost and time. This study analyzes software component selection research studies on methodologies, criteria, and multi-criteria decision-making (MCDM) techniques. The key study findings are: first, comprehensive standardized criteria for software component selection are lacking, with ambiguous terminology used in research. Second, current ad hoc selection processes need streamlining to reduce time, cost, and effort. Thus, an integrated approach is required to aid decision-makers. The review suggests developing automated tools or decision support systems combining multiple criteria decision methods to improve selection accuracy and efficiency. Standardized criteria catalogs can also assist software developers in the evaluation. The findings highlight that despite extensive academic research, component selection in practice remains sub-optimal. By informing future research and tool development, this review can benefit practitioners to systematically select the most appropriate software components meeting software requirements
Classification of endometrial adenocarcinoma using histopathology images with extreme learning machine method
Get PDFAs many as 70-80% of endometrial cancer cases are endometrial adenocarcinoma. Histopathological assessment is based on the degree of differentiation, into well-differentiated, moderate-differentiated, and poorly-differentiated. Management and prognosis differ between grades, so differential diagnosis in determining the degree of tumor differentiation is crucial for appropriate treatment decisions. Histopathological image analysis offers detailed diagnostic results, but manual analysis by a pathologist is very complicated, error-prone, quite tedious, and time-consuming. Therefore, an automatic diagnostic system is needed to assist pathologists in grading the tumor. This research aims to determine the degree of differentiation of endometrial adenocarcinoma based on histopathological images. The extreme learning machine (ELM) method performs image classification with gray level run long matrix (GLRLM) features and a combination of local binary pattern (LBP)-GLRLM features as input. Experimental results show that the ELM model can achieve satisfactory performance. Training accuracy, testing accuracy, and model precision with GLRLM features were 97.13%, 91.33%, and 80% and combined LBPGLRLM features were 91.03%, 71.33%, and 100%. Overall, the model created can determine the degree of tumor differentiation and is useful in providing a second opinion for pathologists
A novel FFNN-AHO hybrid predictive model for enhancing the performance of jet-cooled PVT system
Get PDFPhotovoltaic-thermal (PVT) systems are common in the conversion of solar energy to electrical and thermal energy. The performance of such systems depends on the environmental conditions in which these systems are applied. This paper presents a parametric study of a jet-cooling PVT system with a staggered distribution of the jets. A feedforward neural network (FFNN) is proposed as a novel predictive model for analyzing the characteristics of the PVT system and its thermal and electrical performance. Moreover, a novel optimization algorithm called archerfish hunting optimizer (AHO) is applied to obtain the optimal structure and elements of the proposed FFNN. The PVT system variables considered as inputs to the FFNN-AHO model are flow rate, wind speed, solar irradiance, and ambient temperature. The average temperature of the PV reaches a maximum of 45.84 ºC, and the maximum temperature un-uniformity reaches to 3.59 ºC. The studied PVT system achieved maximum electrical, thermal, and overall efficiencies of 14.23%, 54.43%, and 68.1%, respectively. Moreover, the results demonstrate that the FFNN-AHO hybrid model provides highly accurate PVT system performance prediction. The correlation coefficient between the actual and predicted data is close to 1, indicating a strong correlation and confirming the reliability and effectiveness of the FFNN-AHO model
Efficient deep learning models for Telugu handwritten text recognition
Get PDFOptical character recognition (OCR) technology is indispensable for converting and analyzing text from various sources into a format that is editable and searchable. Telugu handwriting presents notable challenges due to the resemblance of characters, the extensive character set, and the need to segment overlapping characters. To segment the overlapping characters, we assess the width of small characters within a word and segment the overlapping characters accordingly. This method is well suited for the segmentation of overlapping compound characters. To address the recognition of similar characters with less training periods we have used ResNet 18 and SqueezeNet models which have achieved character recognition rates of 95% and 94% respectively
Deep learning based hybrid precoder for optimal power allocation to improve the performance of massive MIMO
Get PDFHybrid precoding is a significant procedure for decreasing the hardware complexity and power usage in massive multiple-input multiple-output (MIMO) systems. However, the effectiveness of hybrid precoding is highly dependent on precise channel state info and designing of the beamforming matrix. In recent years, deep learning-based approaches have emerged as a promising solution to address these challenges. This research focuses on improving the performance of massive MIMO systems. However, several methods have been introduced to develop the hybrid precoding model, but these models suffer from several issues such as complexity, interference and quantization error. Currently, deep learning-based methods have gained huge attention in this domain where these methods learn from the data and try to overcome the challenges. Here, a deep learning-based model is presented where our main aim is to develop a hybrid precoder along with the deep learning-based optimal power allocation model. Therefore, the proposed model overcomes the issue of hybrid precoding and power distribution resulting in improving the overall performance of massive MIMO systems on the parameters such as spectral efficiency (SE) and the sum rate
Improved vigenere using affine functions surrounded by two genetic crossovers for image encryption
Get PDFThis paper presents an improved method for encrypting color images, surpassing the effectiveness of genetic crossover and substitution operations. The technique incorporates dynamic random functions to enhance the integrity of the resulting vector, increasing temporal complexity to thwart potential attacks. The improvement involves integrating genetic crossover and utilizing two extensive pseudorandom replacement tables derived from established chaotic maps in cryptography. Following the controlled vectorization of the original image, our approach initiates with a first genetic crossover inspired by deoxyribonucleic acid (DNA) behavior at the pixel level. This genetic crossover is succeeded by a confusion-diffusion lap, reinforcing the connection between encrypted pixels and their neighboring counterparts. The confusion-diffusion process employs dynamic pseudorandom affine functions at the pixel level. Then a second genetic crossover operator is applied. Simulations conducted on a diverse set of images with varying sizes and formats showcase the robustness of our method against statistical, brute-force, and differential attacks
Notice of Retraction Design of mean filter using field programmable gate arrays for digital images
Get PDFNotice of Retraction-----------------------------------------------------------------------After careful and considered review of the content of this paper by a duly constituted expert committee, this paper has been found to be in violation of IAES's Publication Principles.We hereby retract the content of this paper. Reasonable effort should be made to remove all past references to this paper.The presenting author of this paper has the option to appeal this decision by contacting ijeecs.iaes@gmail.com.-----------------------------------------------------------------------In this paper, we design and analysis of mean filter using field programmable gate arrays (FPGAs) for digital images, FPGAs are integrated circuits consisting of interconnections that connect programmable internal hardware blocks allows users to customize operations for a specific application. FPGA is an ideal choice for real-time image processing, these FPGA devices are controlled in Verilog or VHDL languages, allowing to design at different levels and adapt to design changes or even support new applications throughout the life of the component. Digital image filtering is the most important task in image processing and with the help of computers, image recognition involves identifying and classifying objects in an image. This paper design of mean filter for digital image processing, implementation and analysis of image processing algorithms on FPGAs. The results obtained on the FPGA are compared and analyzed with the results by MATLAB software
Extracting contextual insights from user reviews for recommender systems: a novel method
Get PDFRecommender systems (RS) primarily rely on user feedback as a core foundation for making recommendations. Traditional recommenders predominantly rely on historical data, which often presents challenges due to data scarcity issues. Despite containing a substantial wealth of valuable and comprehensive knowledge, user reviews remain largely overlooked by many existing recommender systems. Within these reviews, there lies an opportunity to extract valuable insights, including user preferences and contextual information, which could be seamlessly integrated into recommender systems to significantly enhance the accuracy of the recommendations they provide. This paper introduces an innovative approach to building context-aware RS, spanning from data extraction to ratings prediction. Our approach revolves around three essential components. The first component involves corpus creation, leveraging Dbpedia as a data source. The second component encompasses a tailored named entity recognition (NER) mechanism for the extraction of contextual data. This NER system harnesses the power of advanced models such as bidirectional encoder representations from transformers (BERT), bidirectional long short term memory (Bi-LSTM), and bidirectional conditional random field (Bi-CRF). The final component introduces a novel variation of factorization machines for the prediction of ratings called contextual factorization machines. Our experimental results showcase robust performance in both the contextual data extraction phase and the ratings prediction phase, surpassing the capabilities of existing state-of-the-art methods. These findings underscore the significant potential of our approach to elevate the quality of recommendations within the realm of context-aware recommender systems
Design and development of arduino-based automation home system using the internet of things
Get PDFThe home automation system described in this paper is low-cost, dependable, and versatile. It uses an Arduino microcontroller and Bluetooth internet protocol (IP) connectivity to allow authorized users to remotely access and control devices. The suggested system employs the internet of things (IoT), which is server-independent, to manage human-desired appliances ranging from industrial machinery to consumer products. In this project, we have taken a Bluetooth module that is programmed through an Arduino Nano to control various devices auto-switching of mechanical devices and monitoring of water level within a range of 130 m using an Android application. This is done to show the effectiveness and viability of this system. Each bulb was switched on/off remotely using a mobile phone successfully. The operation of the water pump attached to the source bucket were controlled from the phone while in manual mode and controlled by an ultrasonic sensor while in automatic mode. It enables remote control of a number of devices, including lights and pumps, and decision-making based on sensor feedback