Статьи журнала - International Journal of Intelligent Systems and Applications
Все статьи: 1159
Load balancing in multicore systems using heuristics based approach
Статья научная
Multicore processing is advantageous over single core processors in the present highly advanced time critical applications. The tasks in real time applications need to be completed within the prescribed deadlines. Based on this philosophy, the proposed paper discusses the concept of load balancing algorithms in such a way that the work load is equally distributed amongst all cores in the processor. The equal distribution of work load amongst all the cores will result in enhanced utilization and increase in computing speed of application with all the deadlines met. In the heuristic based load balanced algorithm (HBLB), the best task from the set of tasks is selected using the feasibility check window and is assigned to the core. The application of HBLB reduces imbalance among the cores and results in lesser migration leading to low migration overhead. By utilizing all the cores of the multicore system, the computing speed of the application increases tremendously which results in the increase in efficiency of the system. The present paper also discusses the improved version of HBLB, known as Improved_Heuristic Based Load Balancing (Improved_HBLB), which focuses on further reducing the imbalance and the number of backtracks as compared to HBLB algorithm. It was observed that Improved_HBLB gives approximately 10% better results over the HBLB algorithm.
Бесплатно
Local Detectors and Descriptors for Object Class Recognition
Статья научная
Local feature detection and description are widely used for object recognition such as augmented reality applications. There have been a number of evaluations and comparisons between feature detectors and descriptors and between their different implementations. Those evaluations are carried out on random sets of image structures. However, feature detectors and descriptors respond differently depending on the image structure. In this paper, we evaluate the overall performance of the most efficient detectors and descriptors in terms of speed and efficiency. The evaluation is carried out on a set of images of different object classes and structures with different geometric and photometric deformations. This evaluation would be useful for detecting the most suitable detector and descriptor for a particular object recognition application. Moreover, multi-object applications such as digilog books could change the detector and descriptor used based on the current object. From the results, it has been observed that some detectors perform better with certain object classes. Differences in performance of the descriptors vary with different image structures.
Бесплатно
Low Complexity Business Status Update Framework: One Touch Approach for Food and Beverage Industry
Статья научная
Today, many individuals are used to dine-out. However, they are unaware of the business operation on that particular moment of the day. Several times, we end up arriving at the restaurant only to find that it is closed/having a break. Hence, we propose a framework for a system of related applications which solves the above problem by being informative regarding the business operability to the customers. Firstly, a trader side framework that allows food stall operators to inform the status and nature of their business to their customers whether they are open for business or not. Secondly, a customer side framework for the food stall operator customers to view restaurant status, menu and to place booking. Mobile applications are developed based on the proposed framework for both trader and customer. And lastly, a website is developed for the general public to view the business status of the stall operators. By being able to inform customers the status of the business, it will provide convenience to many people in our society. Our contribution will be the aforementioned framework as well as mobile apps and website which provides convenience to many people in our society, in terms of reducing time wastage as well as fuel costs to the stall's destination.
Бесплатно
Статья научная
This paper presents the design of low noise amplifier (LNA) at 2.45 GHz and integrated at 0.18 µm RF CMOS process technology. This type of LNA at 2.45 GHz is use in the Bluetooth receiver. The proposed method is useful to optimize noise performance and power gain while maintaining good input and output matching. The amplifier is designed to be used as first stage of a receiver for wireless communication. The main aim of designer is to achieve low noise figure with improved gain with the help of CMOS technology by using single stage n-MOS amplifier. The simulation results show a forward gain of 14.0 dB, a noise-figure of 0.5 dB and stability factor is approximate unity, in which the circuit operates at 14.2 mA drain current with supply voltage of 3.5 V and biasing voltage of 1.5 V.
Бесплатно
MACO-MOTS: modified ant colony optimization for multi objective task scheduling in cloud environment
Статья научная
Cloud computing is the development of distributed computing, parallel computing, and grid computing, or defined as a commercial implementation of such computer science concepts. One of the main issues in a cloud computing environment is Task scheduling (TS). In Cloud task scheduling, many Non deterministic Polynomial time-hard optimization problem, and many meta-heuristic (MH) algorithms have been proposed to solve it. A task scheduler should adapt its scheduling strategy to changing environment and variable tasks. This paper amends a cloud task scheduling policy based on Modified Ant Colony Optimization (MACO) algorithm. The main contribution of recommended method is to minimize makespan and to perform Multi Objective Task Scheduling (MOTS) process by assigning pheromone amount relative to corresponding virtual machine efficiency. MACO algorithm improves the performance of task scheduling by reducing makespan and degree of imbalance comparatively lower than a basic ACO algorithm by its multi-objective and deliberate nature.
Бесплатно
Статья научная
A large number of diverse engineering applications are frequently modeled using different approaches, viz., a differential equation or by a transfer function and state space. All these descriptions provide a great deal of information about the system, such as stability of the system, its step or impulse response, and its frequency response. The present paper addresses different approaches used to derive mathematical models of first and second order system, developing MATLAB script implementation and building a corresponding Simulink model. The dynamic analysis of electric circuit and system using MATLAB/Simulink has been investigated using different approaches for chosen system parameters..
Бесплатно
MCS-MCMC for Optimising Architectures and Weights of Higher Order Neural Networks
Статья научная
The standard method to train the Higher Order Neural Networks (HONN) is the well-known Backpropagation (BP) algorithm. Yet, the current BP algorithm has several limitations including easily stuck into local minima, particularly when dealing with highly non-linear problems and utilise computationally intensive training algorithms. The current BP algorithm is also relying heavily on the initial weight values and other parameters picked. Therefore, in an attempt to overcome the BP drawbacks, we investigate a method called Modified Cuckoo Search-Markov chain Monté Carlo for optimising the weights in HONN and boost the learning process. This method, which lies in the Swarm Intelligence area, is notably successful in optimisation task. We compared the performance with several HONN-based network models and standard Multilayer Perceptron on four (4) time series datasets: Temperature, Ozone, Gold Close Price and Bitcoin Closing Price from various repositories. Simulation results indicate that this swarm-based algorithm outperformed or at least at par with the network models with current BP algorithm in terms of lower error rate.
Бесплатно
MMeMeR: An Algorithm for Clustering Heterogeneous Data using Rough Set Theory
Статья научная
In recent times enumerable number of clustering algorithms have been developed whose main function is to make sets of objects having almost the same features. But due to the presence of categorical data values, these algorithms face a challenge in their implementation. Also some algorithms which are able to take care of categorical data are not able to process uncertainty in the values and so have stability issues. Thus handling categorical data along with uncertainty has been made necessary owing to such difficulties. So, in 2007 MMR algorithm was developed which was based on basic rough set theory. MMeR was proposed in 2009 which surpassed the results of MMR in taking care of categorical data and it could also handle heterogeneous values as well. SDR and SSDR were postulated in 2011 which were able to handle hybrid data. These two showed more accuracy when compared to MMR and MMeR. In this paper, we further make improvements and conceptualize an algorithm, which we call MMeMeR or Min-Mean-Mean-Roughness. It takes care of uncertainty and also handles heterogeneous data. Standard data sets have been used to gauge its effectiveness over the other methods.
Бесплатно
Статья научная
Cloud computing is considered a pattern for distributed and heterogeneous computing derived from many resources, and requests aim to share resources. Recently, cloud computing is graded among the top best technologies globally, which must be scheduled favorably to maximize providers’ profit and improve service quality for their customers. Scheduling specifies how users’ requests are assigned to virtual machines, and it plays a vital role in the efficiency and capability of the system. Its objective is to have a throughput or complete jobs in minimum time and the highest standard. Scheduling jobs in heterogeneous distributed systems is an NP-hard polynomial indecisive problem that is not solvable in polynomial time for real-time scheduling. The time complexity of jobs is growing exponentially, and this problem has a considerable effect on the quality of cloud services and providers’ efficiencies. The optimization of scheduling-related parameters using heuristic and meta-heuristic algorithms can reduce the search space complexity and execution time. This study intends to represent a fitness function to minimize time and cost parameters. The proposed method uses a multi-purposed weighted genetic algorithm that provides six basic parameters: utility, task execution cost, response time, wait time, Makespan, and throughput to provide comprehensive optimization. The proposed approach improved response and wait times, throughput, Makespan, and utility 16, 9, 7, 8 percentages, respectively, by only a one cost unit reduction, which is dispensable. As a result, both providers and users will experience better services. The statistical tests show that the achieved improvement is valid for 94% of experiments.
Бесплатно
Статья научная
Climate change, a significant and lasting alteration in global weather patterns, is profoundly impacting the stability and predictability of global temperature regimes. As the world continues to grapple with the far-reaching effects of climate change, accurate and timely temperature predictions have become pivotal to various sectors, including agriculture, energy, public health and many more. Crucially, precise temperature forecasting assists in developing effective climate change mitigation and adaptation strategies. With the advent of machine learning techniques, we now have powerful tools that can learn from vast climatic datasets and provide improved predictive performance. This study delves into the comparison of three such advanced machine learning models—XGBoost, Support Vector Machine (SVM), and Random Forest—in predicting daily maximum and minimum temperatures using a 45-year dataset of Visakhapatnam airport. Each model was rigorously trained and evaluated based on key performance metrics including training loss, Mean Absolute Error (MAE), Mean Squared Error (MSE), Root Mean Squared Error (RMSE), R2 score, Mean Absolute Percentage Error (MAPE), and Explained Variance Score. Although there was no clear dominance of a single model across all metrics, SVM and Random Forest showed slightly superior performance on several measures. These findings not only highlight the potential of machine learning techniques in enhancing the accuracy of temperature forecasting but also stress the importance of selecting an appropriate model and performance metrics aligned with the requirements of the task at hand. This research accomplishes a thorough comparative analysis, conducts a rigorous evaluation of the models, highlights the significance of model selection.
Бесплатно
Статья научная
Along with the growth of the Internet, social media usage has drastically expanded. As people share their opinions and ideas more frequently on the Internet and through various social media platforms, there has been a notable rise in the number of consumer phrases that contain sentiment data. According to reports, cyberbullying frequently leads to severe emotional and physical suffering, especially in women and young children. In certain instances, it has even been reported that sufferers attempt suicide. The bully may occasionally attempt to destroy any proof they believe to be on their side. Even if the victim gets the evidence, it will still be a long time before they get justice at that point. This work used OCR, NLP, and machine learning to detect cyberbullying in photos in order to design and execute a practical method to recognize cyberbullying from images. Eight classifier techniques are used to compare the accuracy of these algorithms against the BoW Model and the TF-IDF, two key features. These classifiers are used to understand and recognize bullying behaviors. Based on testing the suggested method on the cyberbullying dataset, it was shown that linear SVC after OCR and logistic regression perform better and achieve the best accuracy of 96 percent. This study aid in providing a good outline that shapes the methods for detecting online bullying from a screenshot with design and implementation details.
Бесплатно
Статья научная
Error pattern recognition is a routine job in the military to provide corrective guidelines to the shooter. Errors can be recognized with a visual approach based on the spreading pattern of bullets on the target board, which are categorized into four categories: long horizontal error, long vertical error, bi-focal error, and scattered error. Currently, this process is performed manually and requires active human involvement. Similarly, an automated system to predict the future performance of a shooter is not available in the military domain. Moreover, the performance of a shooter depends on several factors, including age, weather, ammunition type, availability of light, previous scores, shooting range, classification of firing, and other factors. The military domain has not addressed the automatic prediction of such performance. While error correction and performance analysis have been extensively explored in the field of sports, their application within the military domain remains an untapped area of research and investigation. Numerous recent endeavors have suggested the utilization of deep learning to tackle this challenge. However, the absence of real-time data poses a significant obstacle, rendering these solutions seemingly impractical. In this paper, we have applied machine- learning approaches and adopted the best algorithm to automate the error pattern recognition system within a military domain. Our proposed methodology has two modules. The first module uses various algorithms and finds a random forest classifier that can do better to recognize the pattern of error and in the second phase, we used the AdaBoost classifier to predict the score and performance of a firer. Several experiments have been conducted, and the results show an average accuracy of 0.968 using Random Forest to recognize the pattern of error and an accuracy of 0.69 using AdaBoost to predict score performance. The data has been collected from the real-time environment of the military domain and experiments have been carried out using real-time scenarios with the military in mind.
Бесплатно
Managing Lexical Ambiguity in the Generation of Referring Expressions
Статья научная
Most existing algorithms for the Generation of Referring Expressions (GRE) tend to produce distinguishing descriptions at the semantic level, disregarding the ways in which surface issues (e.g. linguistic ambiguity) can affect their quality. In this article, we highlight limitations in an existing GRE algorithm that takes lexical ambiguity into account, and put forward some ideas to address those limitations. The proposed ideas are implemented in a GRE algorithm. We show that the revised algorithm successfully generates optimal referring expressions without greatly increasing the computational complexity of the (original) algorithm.
Бесплатно
Margin Based Learning: A Framework for Acoustic Model Parameter Estimation
Статья научная
Statistical learning theory has been introduced in the field of machine learning since last three decades. In speech recognition application, SLT combines generalization function and empirical risk in single margin based objective function for optimization. This paper incorporated separation (misclassification) measures conforming to conventional discriminative training criterion in loss function definition of margin based method to derive the mathematical framework for acoustic model parameter estimation and discuss some important issues related to hinge loss function of the derived model to enhance the performance of speech recognition system.
Бесплатно
Статья научная
We analyzed the dynamic behavior of the damping system with a two-mass damper pendulum. The equations of motion of nonlinear systems were built. AFC equation systems have been identified in the linear formulation. Proposed and implemented a new numerical method of determining the optimum parameters of optimal settings two-mass damper.
Бесплатно
Mathematical model of the damping process in a one system with a ball vibration absorber
Статья научная
The forced oscillations of the damping mechanical system of solids "Ball Vibration Absorber (BVA) with linearly viscous resistance – a movable carrier body" under the influence of external harmonic excitation are considered. Based on Appell's formalism, the dynamic equations for the joint motion of a heavy ball without sliding into a spherical cavity of a carrier body are formulated and numerically studied. The amplitude-frequency characteristic of the damping mechanical system and the curves of the dependences of the maximum amplitude of the oscillations of the carrier body on the values of the radius of the spherical cavity and the coefficient of viscous resistance of the BVA are obtained. The conditions and restrictions on the rolling of a heavy ball in the spherical recess of the absorber without sliding are determined.
Бесплатно
Mathematical model of the dynamics in a one nonholonomic vibration protection system
Статья научная
Dynamic behavior of a heavy homogeneous sphere in a spherical cavity of a supporting body that performs specified translational movements in space has been studied. Using the Appel formalism, the equations of ball motion in a moving spherical cavity without slip are constructed and a numerical analysis of the evolution of the ball motion is carried out.
Бесплатно
Measuring and Evaluation on Priority Lanes
Статья научная
Along with economic development, cities are increasingly congested in China. In order to eliminate peak-hour congestion, many cities establish priority lanes, commonly bus lanes. Although priority lanes could help Local Authorities gain its short-term management objectives, at the same time, it would greatly infringe on the legitimate rights of other vehicles and waste the scarce road resources, which is rigorously proved by mathematical models in this paper. In the long run, priority lanes would make social conflicts more intensified, and therefore highly undesirable. On the contrary, the social system engineering, combined with High Occupancy Vehicle (HOV) lanes and High Occupancy Toll (HOT) lanes, is the right way to alleviate overcrowding and build a Low-Carbon harmonious society.
Бесплатно
Mechatronics Design of a Mobile Robot System
Статья научная
Mobile robot motion control is simplified to a DC motor motion control that may include gear system. The simplest and widespread approach to control the mobile robot motion is the differential drive style, it consists of two in-lines with each a DC motor. Both DC motors are independently powered so the desired movements will rely on how these two DC motors are commanded. Thedevelop design, model and control of Mechatronics mobile robotic system is presented in this paper. The developed robotic system is intended for research purposes as well as for educational process. The model of proposed mobile robot was created and verified using MATLAB-Simulink software.
Бесплатно
MediBERT: A Medical Chatbot Built Using KeyBERT, BioBERT and GPT-2
Статья научная
The emergence of chatbots over the last 50 years has been the primary consequence of the need of a virtual aid. Unlike their biological anthropomorphic counterpart in the form of fellow homo sapiens, chatbots have the ability to instantaneously present themselves at the user's need and convenience. Be it for something as benign as feeling the need of a friend to talk to, to a more dire case such as medical assistance, chatbots are unequivocally ubiquitous in their utility. This paper aims to develop one such chatbot that is capable of not only analyzing human text (and speech in the near future), but also refining the ability to assist them medically through the process of accumulating data from relevant datasets. Although Recurrent Neural Networks (RNNs) are often used to develop chatbots, the constant presence of the vanishing gradient issue brought about by backpropagation, coupled with the cumbersome process of sequentially parsing each word individually has led to the increased usage of Transformer Neural Networks (TNNs) instead, which parses entire sentences at once while simultaneously giving context to it via embeddings, leading to increased parallelization. Two variants of the TNN Bidirectional Encoder Representations from Transformers (BERT), namely KeyBERT and BioBERT, are used for tagging the keywords in each sentence and for contextual vectorization into Q/A pairs for matrix multiplication, respectively. A final layer of GPT-2 (Generative Pre-trained Transformer) is applied to fine-tune the results from the BioBERT into a form that is human readable. The outcome of such an attempt could potentially lessen the need for trips to the nearest physician, and the temporal delay and financial resources required to do so.
Бесплатно
