International Journal of Intelligent Systems and Applications @ijisa
Journal articles - International Journal of Intelligent Systems and Applications
All articles: 1264
CervixCan-Net: An Enhanced Cervical Cancer Classification Approach using Deep Learning
Scientific article
One of the biggest causes of cancer-related fatalities among women is still Cervical cancer, especially in low and middle-income nations where access to broad screening and early detection may be limited. Cervical cancer is curable if detected in its early stages, but asymptomatic progression frequently results in late diagnosis, which makes treatment more difficult and lowers survival chances. Even though they work well, current screening methods including liquid-based cytology and Pap smears have drawbacks in terms of consistency, sensitivity, and specificity. Recent developments in Deep Learning and Artificial Intelligence have shown promise for greatly improving Cervical cancer detection and diagnosis. In this work, we have introduced CervixCan-Net, a novel Deep Learning based model created for the precise classification of Cervical cancer from histopathology images. Our approach offers a solid and dependable classification solution by addressing common problems like overfitting and computational inefficiency. CervixCan-Net performs better than many state-of-the-art models according to a comparison investigation. CervixCan-Net, with an impressive test accuracy of 99.83%, provides a scalable, automated Cervical cancer classification solution that has great promise for improving patient outcomes and diagnostic accuracy.
Free
Challenges with Sentiment Analysis of On-line Micro-texts
Scientific article
With the evolution of World Wide Web (WWW) 2.0 and the emergence of many micro-blogging and social networking sites like Twitter, the internet has become a massive source of short textual messages called on-line micro-texts, which are limited to a few number of characters (e.g. 140 characters on Twitter). These on-line micro-texts are considered as real-time text streams. On-line micro-texts are extremely subjective; they contain opinions about various events, social issues, personalities, and products. However, despite being so voluminous in quantity, the qualitative nature of these micro-texts is very inconsistent. These qualitative inconsistencies of raw on-line micro-texts impose many challenges in sentiment analysis of on-line micro-texts by using the established methods of sentiment analysis of unstructured reviews. This paper presents many challenges and issues observed during sentiment analysis of On-line Micro-texts.
Free
Chaotic Genetic Algorithm based on Lorenz Chaotic System for Optimization Problems
Scientific article
Very recently evolutionary optimization algorithms use the Genetic Algorithm to improve the result of Optimization problems. Several processes of the Genetic Algorithm are based on 'Random', that is fundamental to evolutionary algorithms, but important defections in the Genetic Algorithm are local convergence and high tolerances in the results, they have happened for randomness reason. In this paper we have prepared pseudo random numbers by Lorenz chaotic system for operators of Genetic Algorithm to avoid local convergence. The experimental results show that the proposed method is much more efficient in comparison with the traditional Genetic Algorithm for solving optimization problems.
Free
Scientific article
Transient analysis is carried out for the single-phase grounding fault in small current grounding system, the transient grounding current expression is derived, and the influence factors are analyzed. Introduces a method for non-stationary and non-linear signal analysis method –Hilbert Huang transform (HHT) to analyze the single phase grounding fault in small current grounding system, HHT can be better used to extract the abundant transient time frequency information from the non-stationary and nonlinear fault current signals. The empirical mode decomposition (EMD) process and the normalized Hilbert Huang transform (NHHT) algorithm are presented, NHHT is used to analyze and verify an example of the nonlinear and non-stationary amplitude modulation signals. Build a small current grounding system in the EMTP_ATP environment, by selecting the appropriate time window to extract the transient signals, NHHT is used to analyze the transient current signals, and the Hilbert amplitude spectrum and the Hilbert marginal spectrum of the zero sequence transient current signals are obtained. Finally, the influences of the fault phase and the grounding resistance on the time-frequency characteristics of the signals are analyzed.
Free
Classification and regression trees (CART) for predictive modeling in blended learning
Scientific article
Today, Internet and Web technologies not only provide students opportunities for flexible interactivity with study materials, peers and instructors, but also generate large amounts of usage data that can be processed and reveal behavioral patterns of study and learning. This study analyzed data extracted from a Moodle-based blended learning course, to build a student model that predicts course performance. CART decision tree algorithm was used to classify students and predict those at risk, based on the impact of four online activities: message exchanging, group wiki content creation, course files opening and online quiz taking. The overall percentage of correct classifications was about 99.1%, proving the model sensitive to identify very specific groups at risk.
Free
Classification of EEG signals using Hyperbolic Tangent-Tangent Plot
Scientific article
In this paper, a novel signal processing method is suggested for classifying epileptic seizures. To this end, first the Tangent and Hyperbolic Tangent of signals are calculated and then are classified into two classes: normal (or interictal) and ictal, using a proposed classifier. The results of this method show that the classification accuracy of normal and ictal classes (97.41%) has been higher than interictal and ictal classes (92.83%) and generally, it has a good potential to become a useful tool for physicians.
Free
Classification of Forest Land Information Using Environment Satellite (HJ-1) Data
Scientific article
For researching properties of HJ-1A CCD camera multi-spectral data in performance on extraction of land features information, this paper selected the east area of NiLeke forest farm in the western Tianshan mountain as the study area, and analyzed different accuracies for HJ-1A CCD data in identifying forest land categories using various classification methods. Firstly, maximum-likelihood classifier, Mahalanobis distance classifier, minimum distance classifier and K-means classifier were used to category land use types with two different scales on HJ-1A CCD1 and Landsat5 TM images, and analyzed separately with confusion matrix. Secondly, forest land types were distinguished by texture information and the smallest polygon size using K-NN method based on clustering algorithm. The comparing results show: at first, different classification system have different accuracy. In the first land use classification system, the accuracy of HJ-1A CCD1 images are lower than TM images, but higher in the second land use classification system. Secondly, accuracy result of maximum-likelihood classification is the best method to classify land use types. In the first land use classification system, TM total accuracy is up to 85.1% and Kappa coefficient is 0.8. In the second land use classification system, the result is up to 85.4% and kappa coefficient is 0.74.Thirdly, judgment both from the view of visual interpretation and quantitative accuracy testes, non-supervised method with K-means classifier has low qualities where many land features have characters of scattered distribution and small different spectrum information. Finally, the experiment proved that there were good vector results on HJ-1A remote sensing image in the view of visual judgment, and extracted deferent forest land by the overall accuracy 87% with the supports by those variables’ distribution knowledge, such as conifer, mixed forest, broadleaf, shrubby.
Free
Classification of Heart Rate Signals during Meditation using Lyapunov Exponents and Entropy
Scientific article
Meditation is commonly perceived as an alternative medicine method of psychological diseases management tool that assist in alleviating depression and anxiety disorders. The purpose of this study is to evaluate the accuracy of different classifiers on the heart rate signals in a specific psychological state. Two types of heart rate time series (before, and during meditation) of 25 healthy women are collected in the meditation clinic in Mashhad. Nonlinear features such as Lyapunov Exponents and Entropy were extracted. To evaluate performance of the classifiers, the classification accuracies and mean square error (MSE) of the classifiers were examined. Different classifiers were tested and the studies confirmed that for the heart rate signals, Quadratic classifier trained on Lyapunov Exponents and Entropy results in higher classification accuracy. The classification accuracy of the Quadratic classifier is 92.31%. However, the accuracies of Fisher and k-Nearest Neighbor (k-NN) classifiers are encouraging. The classification results demonstrate that the dynamical measures are useful parameters which contain comprehensive information about signals and the Quadratic classifier using nonlinear features can be useful in analyzing the heart rate signals in a specific psychological state.
Free
Classification of Images of Skin Lesion Using Deep Learning
Scientific article
Skin cancer is among common and rapidly increasing human malignancies, which can be diagnosed visually. The diagnosis begins with preliminary medical screening and by dermoscopic examination, histopathological examination, and proceeding to the biopsy. This screening and diagnosis can be automated using machine learning tools and techniques. Artificial neural networks are helping a lot in medical diagnosis applications. In this research, skin images are classified into 7 different classes of skin cancer using deep learning methodology, then analyzed the results w.r.t to their respective precision, recall, support, and accuracy to find its practical applicability. This model is efficient in comparison to the detection of skin cancer with human eyes. Human eyes detection can be 79% accurate at most. Thus, having a scientific method of diagnosis can help the doctors and practitioners to accurately identify the cancer and its type. The model provides 80% accuracy on average for all 7 types of skin diseases, thus being more reliable than human eye examination. It will help the doctors to diagnose the skin diseases more confidently. The model has only 2 misclassified predictions for Basal cell carcinoma and Vascular lesions. However, Actinic keratosis diagnosis is most accurately predicted.
Free
Classification of Medicinal Plant Leaves using Deep Learning Algorithms
Scientific article
This research explores the automated leaf-based identification of medicinal plants, utilizing machine learning and deep learning techniques to address the crucial need for efficient plant classification. Driven by the vast potential of medicinal plants in pharmaceutical development and healthcare, the study aims to surpass the limitations of existing methodologies through thorough experimentation and comparative analysis. The primary goal is to develop a robust and automated solution for classifying medicinal plants based on leaf morphology. The methodology encompasses acquiring diverse datasets. Specifically, Set 1 data is processed by applying resizing, rescaling, saturation adjustment, and noise removal, while Set 2 data is processed by applying resizing, rescaling, saturation adjustment, noise removal, and PCA (Principal Component Analysis). The proposed algorithms include Support Vector Machines (SVM), Convolutional Neural Networks (CNNs), YOLOv8, Vision Transformer (ViT), ResNet, and Artificial Neural Networks (ANN). The study evaluates the efficacy and effectiveness of each algorithm in plant classification using metrics such as accuracy, recall, precision, and F1 score. Notably, the ResNet model achieved 93.8% and 94.8% accuracy in Set 1 and Set 2, respectively. The SVM model demonstrated 56.5% and 56.6% accuracy in Set 1 and Set 2, while the Vision Transformer (ViT) model achieved 84.9% and 74.4% accuracy in Set 1 and Set 2, respectively. The CNN model showcased high accuracy at 96.7% and 94.8% in Set 1 and Set 2, followed closely by the ANN model with 96.7% and 96.6% accuracy. Lastly, the YOLOv8 model achieved 96.0% and 95.1% accuracy in Set 1 and Set 2, respectively. The comparative analysis identifies CNN and ANN as the top-performing algorithms. This research significantly contributes to the advancement of medicinal plant identification, pharmaceutical research, and environmental conservation efforts, emphasizing the potential of deep learning techniques in addressing complex classification tasks.
Free
Classification of Reusable Components Based on Clustering
Scientific article
Software reuse is the process of implementing or updating software systems using existing software components. A good software reuse process facilitates the increase of productivity, quality and reliability. It decreases the cost and implementation time as compared to develop new system. Despite its many benefits we cannot achieve its full benefits. The reason behind this is that software reuse is often done in an informal and haphazard way. If done systematically, then we can achieve its full benefits. This research proposes a method through which we will classify the reusable components in proper way to get the full benefits of reusability. We classify the reusable components according to their clusters. Clusters are made on the basis of parameters provided with components. We develop an algorithm for assigning clusters to the reusable components.
Free
Cloud Task Scheduling for Load Balancing based on Intelligent Strategy
Scientific article
Cloud computing is a type of parallel and distributed system consisting of a collection of interconnected and virtual computers. With the increasing demand and benefits of cloud computing infrastructure, different computing can be performed on cloud environment. One of the fundamental issues in this environment is related to task scheduling. Cloud task scheduling is an NP-hard optimization problem, and many meta-heuristic algorithms have been proposed to solve it. A good task scheduler should adapt its scheduling strategy to the changing environment and the types of tasks. In this paper a cloud task scheduling policy based on ant colony optimization algorithm for load balancing compared with different scheduling algorithms has been proposed. Ant Colony Optimization (ACO) is random optimization search approach that will be used for allocating the incoming jobs to the virtual machines. The main contribution of our work is to balance the system load while trying to minimizing the make span of a given tasks set. The load balancing factor, related to the job finishing rate, is proposed to make the job finishing rate at different resource being similar and the ability of the load balancing will be improved. The proposed scheduling strategy was simulated using Cloudsim toolkit package. Experimental results showed that, the proposed algorithm outperformed scheduling algorithms that are based on the basic ACO or Modified Ant Colony Optimization (MACO).
Free
Cloud Theory and Fractal Application in Virtual Plants
Scientific article
Plants is an important component of natural scene. Unfortunately, due to high level complexity of the structure of plant, simulating plant becomes extremely a difficult task. When the fractal theory is imported, it provides a broader development space for the plant modeling. With the development of the fractal research, virtual plant has become a hot and interesting research topic in computer graphics area. The virtual plants technology is very important in guiding the crop production, implementing the agriculture informationization and constructing the virtual environment. At present a single virtual plant modeling technology is quite mature, the method to generate a body of plants often uses the even algorithm or the normal algorithm, but a body of plants in the real world is not even, and is not normal also, the cloud model relaxes the precise determination membership function to expectation function with normal distributed membership degree, combines ambiguity and randomness organically to fit the real world objectively. So it has general applicability, producing a body of plants based on the cloud model can simulate plant's condition and the distribution well.
Free
Clustering matrix sequences based on the iterative dynamic time deformation procedure
Scientific article
The techniques of Dynamic Time Warping (DTW) have shown a great efficiency for clustering time series. On the other hand, it may lead to sufficiently high computational loads when it comes to processing long data sequences. For this reason, it may be appropriate to develop an iterative DTW procedure to be capable of shrinking time sequences. And later on, a clustering approach is proposed for the previously reduced data (by means of the iterative DTW). Experimental modeling tests were performed for proving its efficiency.
Free
Scientific article
Clustering is the process of grouping a set of data objects into multiple groups or clusters with high similarities and dissimilarities. Dissimilarities and Similarities are assessed on the attribute values describing the objects and often involve distance measures. Clustering acts as a data mining tool by having its roots in many application areas such as biology, security, business intelligence, web search etc. Our Institute is currently using a software application with a name "Merit System", which evaluates the performance of the staff members regarding their level of teaching by considering various factors. It computes the performance level by collecting feedback from every student. It gives the appraisal result in the form of 30 points earned to every staff member. It acts as a tool for the management of our college to gauge the performance level of the teacher which in turn helps them in assessing annual increments and other promotions. The main drawback of this system is its inability in grouping of staff members like Group-A, Group-B, Group-C etc. Because, many of the staff members have scored the performance points in the range of 21 to 30 which will creates lot of ambiguities to the management to make clusters of staff members to these groups. This issue is the prime concern of this paper and it was given with an approach to solve this problem by considering possible optimum soft computing technique that includes Feed Forward Neural Network approach.
Free
Scientific article
In this paper, use the Taguchi experiment design method for achieving optimal design to reduce cogging torque and torque ripple of a surface permanent magnet motor. Cogging torque is one cause of vibration and noise, so reducing cogging torque is an important issue Results from simulations, indicate reduction of cogging torque and torque ripple while increasing the average of motor’s output torque. How do simulations and achieve the optimum design, by using the Taguchi method and simulations by using the finite element method has been done.
Free
Cognitive Agents and Learning Problems
Scientific article
Goals, Operators, Methods, and Selection rules (GOMS) model is a widely recognised concept in Human-Computer Interaction (HCI). Since the initial idea, several GOMS techniques were developed that were used for analysis, differing in their form defined by the logical structure and prediction power. Through defined operators and methods and following the certain rules, the user can reach a specific goal. This work represents an effort to apply GOMS method in the field of artificial intelligence, specifically on a state-space search problems. Card, Morgan, Newman GOMS (CMN-GOMS) model has been chosen, since it represents ground-floor of the GOMS idea that solves the given task through a sequence of operators. Compared with the informed search algorithms for solving the given task, CMN-GOMS model gave better results. Moreover, it was shown that this model could be used in any other space motion problem in the natural environment. LEGO® MINDSTORMS® EV3 robot was used to demonstrate the application of GOMS model in real world pathfinding problems and as a test-bed for comparing proposed model with well-known search algorithms.
Free
Collaborative E-Learning Process Discovery in Multi-tenant Cloud
Scientific article
With the appearance of the COVID-19 pandemic, the practice of e-learning in the cloud makes it possible to: avoid the problem of overloading the institutions infrastructure resources, manage a large number of learners and improve collaboration and synchronous learning. In this paper, we propose a new e-leaning process management approach in cloud named CLP-in-Cloud (for Collaborative Learning Process in Cloud). CLP-in-Cloud is composed of two steps: i) design general, configurable and multi-tenant e-Learning Process as a Service (LPaaS) that meets different needs of institutions. ii) to fulfill the user needs, developpe a functional and non-functional awareness LPaaS discovery module. For functional needs, we adopt the algorithm A* and for non-functional needs we adopt a linear programming algorithm. Our developed system allows learners to discover and search their preferred configurable learning process in a multi-tenancy Cloud architecture. In order to help to discover interesting process, we come up with a recommendation module. Experimentations proved that our system is effective in reducing the execution time and in finding appropriate results for the user request.
Free
Collision-free Random Paths between Two Points
Scientific article
This paper proposes a collision-free path planning algorithm based on the generation of random paths between two points. The proposed work applies to many fields such as education, economics, computer science and AI, military, and other fields of applied sciences. Our work has spanned several phases, where in the first phase a novel computer algorithm to generate random paths between two points in space has been developed. The aim was to be able to generate paths between two points in real-time that cannot be predicted in advance. In the second phase, we have developed an ontology that describes the domain of discourse. The aim was two folds; firstly, to provide an optimized generation of best points that are closer to the target point. Secondly, to provide sharable, reusable ontological objects that can be deployed to other projects. We reinforced our solution by the initiation of several case studies that have been designed using and extending our work. One problem that we have faced in some cases is the existence of some obstacles between the starting and the ending point. For example, in our work towards the automation of a navigation system for drones, we faced some obstacles like trees, no flying zones, and buildings. This problem is also applicable to mobile robots and other unmanned vehicles, where fee-collision mobility is necessary. In this phase, we have reworked the algorithm to generate random paths between two points P0(x0, y0), Pn(xn, yn) with obstacles. Our generated random paths are placed within circles that are centered in Pn: c1, c2, …, cn-1, which passes thru the points P1, P2, …, Pn-1 respectively. Point Pi may approach Pn if it takes any position within circle c centered in Pn with radius PiPn and satisfies some constraints, discussed in detail in the paper, which insure that the selected paths do not fall within obstacles and reach the target point. we also classified the generated paths based on given properties such as the longest path, shortest path, and paths with some given costs. The resulted algorithms were very encouraging and leading to the applicability of real-life cases.
Free
Colonial Competitive Optimization Sliding Mode Controller with Application to Robot Manipulator
Scientific article
One of the best nonlinear robust controllers which can be used in uncertain nonlinear systems is sliding mode controller (SMC), but pure SMC results in chattering in a noisy environment. This effect can be eliminated by optimizing the sliding surface slope. This paper investigates a novel methodology in designing a SMC by a new heuristic search, so called "colonial competitive algorithm "in order to tune the sliding surface slope and the switching gain of the discontinuous part in SMC structure. This process decreases the integral of absolute errors which results in tracking the desired inputs by the outputs in designing a controller for robot manipulator. Simulation results prove that the optimized performance obtained through CCA significantly reduces the chattering phenomena and results in better trajectory tracking compared to typical trial and error methods.
Free
