Search

Purpose: Selection of the best stocks for the portfolio as well as allocating the optimal amount of capital per stock in the portfolio are serious challenges in investing in the stock market. The use of machine learning capacities in the process of optimal capital allocation among portfolio assets has received less attention and usually, the same weight is assigned to portfolio stocks or traditional risk assessment methods are used to divide capital between portfolio stocks. The common disadvantage of these methods is that they all use simple and inflexible mechanisms to estimate the performance of a set. The purpose of this paper is to show for the first time, that machine learning can be used to create a more effective mechanism for estimating performance, which leads to a more efficient allocation of capital to portfolio stocks. Methodology: Our proposed framework, uses two predictive models based on machine learning. In the first step, stocks historical information is used in a return forecasting model, then based on the predicted returns, the appropriate stocks of the portfolio are selected. In the second step, a separate forecasting model predicts portfolio returns by taking into account both the forecasted returns in the first model and the expected risk of the stocks. At the end based on the predicted return of the numerous random portfolios, the appropriate weight for each asset is selected. Findings: Comparing the returns of adjusted portfolios with this model and adjusted portfolios with other portfolio optimization methods shows the superiority of the proposed model. Originality/Value: In this paper, by using machine learning models, the process of selecting the appropriate stock of the portfolio and allocating capital among the candidate stocks is done optimally. [ABSTRACT FROM AUTHOR]

Continuous changing in house prices very rapid and observed in last years which depends on many factors like position, area, population in the city and another requirement for individual house price, there is a lot of articles studied and adopted traditional ways for machine learning in house price prediction accurately but few of them focused on individual performance for complicated models. To explore effected factors in prediction way in this paper we applied traditional machine learning theory to investigate many models due to many limited we take some of techniques like (random forest, Extreme Gradient Boosting and stacked generalization model) and applied on several models to obtain more accurate results for house price prediction. We take Baghdad as a sample for house price and impact factor on difference cities, services, population, area, building age and others we mentioned in our paper by taking and applying machine learning house price prediction. [ABSTRACT FROM AUTHOR]

Copyright of Scientific Journal of King Faisal University, Humanities & Management Sciences is the property of Association of Arab Universities and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Internet of Thing (IoT)" is a major and emerging technology where different devices are associated together to operate smartly deprived of human’s intervention. IoT has a great effect on economic, social and commercial life. The extremely large number of associated devices together through different protocols make it vulnerable to be threatened by several types of attacks, such as; Unauthorized Access and Authentication Attacks, Network Attacks, Data Privacy and Integrity Attacks, and Supply Chain Attacks. Building and developing lightweight authentication systems for IoT devices are indeed crucial due to several compelling reasons, including; Resource Constraints, Security Concerns, Scalability, Bandwidth Efficiency, Latency Considerations, Cost Reduction, Long Lifecycle and the overall User Experience. Blockchain and “Machine Learning (ML)” are emerging technologies that may be exploited in solving different security problems. This paper presents a survey of lightweight authentication protocols and schemes that are adopted and applied for IoT environment. The security issues in IoT environment are also discussed besides the different types of attacks that may face the network. The paper held a comparison between the selected works and studies in terms of different criteria including the benefits, the results, the used applications and technologies or methods.

Copyright of Umm Al-Qura University Journal for Languages & Literature is the property of Association of Arab Universities and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Financial Management Perspective / Chashm/&āz-i Mudīriyyat-i Mālī is the property of Shahid Beheshti University and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The use of Deep Learning (DL) for identifying false or misleading information, known as fake news, is a growing area of research. Deep learning, a form of machine learning that utilizes algorithms to learn from large data sets, has shown promise in detecting fake news. The spread of fake news can cause significant harm to society economically, politically, and socially, and it has become increasingly important to find ways to detect and stop its spread. This paper examines current studies that use deep learning methods, such as Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs), as well as the multi-model approach, to detect fake news. It also looks at the use of word embedding models to convert text to vector representations and the datasets used for training models. Furthermore, the paper discusses the use of the attention mechanism in conjunction with deep learning to process sequential data.

Yearly, an amount of funds is spent to achieve defect detection in the principle of infrastructure represented by roads, bridges, and buildings. Urban infrastructure is affected by weather conditions, the natural disasters such as floods and earthquakes. As well as, mistakes during street paving operations and quality of paving materials. Various types of damage may appear in the form of small or vast cracks, which gradually spread, destroying the structure. Therefore, it requires building automatic systems for these inspection operations to guarantee its effectiveness and dependability. Hybrid image processing and machine learning approaches are being applied to guarantee better enhancement outcomes and strength in crack detection. This paper aims to offer a review of road image crack detection techniques that apply image processing with/without machine learning. A total of 32 research articles have been composed and studied for the review which has been issued in publications and conferences in the past years. This research manners a thorough analysis and comparison of various methods to identify the most promising automated methods for crack detection. After analyzing and reviewing previous research using digital image processing methods, it is clear from the results obtained that the best of them is the Franji filter method, whose accuracy is close to 98.7%. While discussing and presenting machine learning techniques and convolutional networks, the deduced results that the best of them is the Support Vector Machine (SVM) technique, whose precision is approximately 98.29%

Generative models try to obtain a probability distribution that is similar to that of observed data. Two different solutions have been proposed in this regard in recent years: one is to minimize the divergence (distance) between the two distributions by maximizing the variational lower bound, and the other is to implicitly reduce the distance between the two distributions through adversarial processes. One of the problems in generative adversarial networks (GANs) is the mode collapse. Mode collapse is a phenomenon in which, for various inputs, the generative model generates low variety or similar images. This paper tries to provide a solution to the mode collapse problem proposing a novel method called variational generative adversarial networks (VGANs). This method exploits variational autoencoders to initialize GANs. In other words, in addition to maximizing the variational lower bound, it also implicitly reduces the distance between the two distributions. Experimental results show that this method can deal with the mode collapse problem better than the state-of-the-art. Moreover, in the qualitative analysis, according to a survey of 136 people on the authenticity of the generated images, the proposed method can generate images more similar to real ones. [ABSTRACT FROM AUTHOR]

Copyright of Iranian Journal of Information Processing & Management is the property of Iranian Information & Documentation Center (IRANDOC) and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Journal of Education & Science is the property of Republic of Iraq Ministry of Higher Education & Scientific Research (MOHESR) and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

In recent decades, various pollutants have been released into the environment in substantial amounts due to human activities. Oil pollution is one of these pollutants and can be harmful to both humans and the environment, making its detection and cleanup highly important. Considering the applicability of the electronic nose methods in various domains, including the food industry, environment, etc, this research aims to measure the feasibility of using electronic nose technology to detect soil contaminated with crude oil. An electronic nose system with eight metal oxide semiconductor sensors was used in this research. The soil used in this research was from one of the farms surrounding Karaj province, which, in the laboratory of the University of Tehran, was contaminated with crude oil at four concentrations: 200 ppm, 500 ppm, 1000 ppm, and 5000 ppm, and then the samples evaluated. This system was examined along with pattern recognition methods such as principal component analysis (PCA), linear discriminant analysis (LDA), and support vector machine (SVM). The results from the pattern recognition methods showed an accuracy of 92% for principal component analysis, 100% for linear discriminant analysis, and 100% for the support vector machine method. Moreover, by employing statistical data analysis using a factorial experiment method within a randomized block design and using decision tree regression, support vector regression, gradient boosted regression, and random forest regression, which has the pollutant detection rate with an average accuracy of 95.5% for the MQ3 sensor, showed the best performance, and detect the difference between the contaminated concentrations. In conclusion, by comparing the results of selected electronic nose sensors with the results obtained by the gas chromatography system on the soil contaminated with crude oil, they have had an acceptable sensitivity. [ABSTRACT FROM AUTHOR]

Copyright of Business Intelligence Management Studies is the property of Allameh Tabatabai University and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Artificial intelligence (AI) applied to accounting and auditing represents a matter that is getting the attention of researchers. This paper provides an understanding of the evolution of AI and its role in shaping accounting information systems (AISs). It aims to summarize the current state of the literature that deals with the impact of AI on AIS, highlighting related aspects, such as the benefits and challenges of AI in AIS research. This review suggests that current studies dealing with this matter are scarce and invites future research to examine how bias and transparency are handled in the context of AI auditing systems and whether human auditors will fully rely on AI outcomes.

Most of the applications used on the internet are Web-Based Applications, that accept critical information from users and store this information in databases. Being connected to the internet, they are susceptible to all kinds of information security threats, including SQL injection attack. SQL injection attacks, and web-based attacks fall in general under the top ten vulnerabilities according to the assessment of the most important information security centers and international networks, such as (OWASP) and (ENSIA), which means they continue to be a major issue in the cyber security field. This paper proposes a method for SQL injection attack detection by using natural language processing techniques (BOW, TF-IDF, Word2Vec, Doc2Vec), and machine learning algorithms (LR, MLP) that allow the machine to automatically learn and detect the characteristic patterns of the query used in SQL injection attacks, which could greatly put an end to attackers’ intervention and provide an appropriate defense mechanism against this type of widespread attack.

Melanoma is considered a serious health disease and one of the most dangerous and deadly types of skin cancer, due to its unlimited spread. Therefore, detection of this disease must be early and sound due to the high mortality rate. It is driven by researchers' desire to use computers to obtain accurate diagnostic systems to help diagnose and detect this disease early. Given the growing interest in cancer prediction, we have presented this paper, a systematic review of recent developments, using artificial intelligence focusing on melanoma skin lesion detection, particularly systems designed on neural networks. Using the neural networks for melanoma detection could be part of system of assistance for dermatologists who must make the final decision on whether to recommend a biopsy if at least one of the dermatologist's diagnoses and the support system (a helpful method) indicate melanoma or to investigate if another type of cancerous lesion exists. In the latter situation, the system can be trained to recognize distinct types of cancerous skin lesions. On the other hand, the system is incapable of making final decisions. Given neural networks' evolutionary patterns, updated, changed, and integrated networks are expected to increase the performance of such systems. Based on the decision fusion, theoretical and applied contributions were studied using traditional classification algorithms and multiple neural networks. The period 2018-2021 has been focused on new trends. Also for the detection of melanomas, the most popular datasets and how they're being used to train neural network models were presented. Furthermore, the field of research emphasized in order to promote better the subject during different directions. Finally, a research agenda was highlighted to advance the field towards the new trends.

Predicting heart attacks using machine learning is an important topic. Medical data sets contain different features, some of which are related to the target group for prediction and some are not. In addition, the data sets are excessively unbalanced, which leads to the bias of machine learning models when modeling heart attacks. To model the unbalanced heart attack data set, this paper proposes the hybridization of Particle swarm optimization (PSO), BAT, and Cuckoo Search (CS) to select the features and adopt the precision for minority classes as a fitness function for each swarm to select the influential features. In order to model the data, set in which the features were selected, it was proposed to use the boosting (Catboost) as a classifier for predicting heart attacks. The proposed method to select features has been compared with each of the three swarms, and the Catboost algorithm has been compared to traditional classification algorithms (naive Bayes, decision trees). The study found that the proposed method of hybridization of the results of the (PSO, BAT, and BCS) algorithms in selecting features is a promising solution in the field of selecting features and increases the accuracy of the system, and that traditional machine learning models are biased in the case of unbalanced data sets and that selecting the important features according to the target class has an impact on the performance of the models, In addition, the definition of hyperparameters reduces the bias of the selected model. The final model achieved an overall accuracy of 96% on the Accuracy scale and 56% on the Precision scale for the minority class

Retinal image analysis is crucial for The classification of retinal diseases such as “Age Related Macular Degeneration (AMD)”, “Diabetic Retinopathy (DR)”, “Retinoblastoma”, “Macular Bunker”, “Retinitis Pigmentosa”, and “Retinal Detachment”. The early detection of such diseases is important insofar as it contributes in mitigating future implications. Many approaches have been developed in the literature for auto-detecting of retinal landmarks and pathologies. The current revolution in deep learning techniques has opened the horizon for researchers in the field of ophthalmology. This paper is a comprehensive review of the deep learning techniques applied for the classification of retinal images, pathology, retinal landmarks, and disease classification. This review is based on indicators such as sensitivity, Area under ROC curve, specificity, F score, and accuracy.

The Internet of Things (IoT) is generally acknowledged as a dramatic change spearheaded by scientists and business executives. The IoT has the potential to improve our daily lives by connecting smart devices to the internet. Due to the limited resources and distant deployment of these IoT devices, securing them is a significant challenge today. This paper focuses primarily on mitigating threats and attacks on realistic artificial intelligence, such as network architecture for smart devices. The text on mitigating attacks in networks, especially those involving mobile nodes, is discussed. We develop and test a new countermeasure against all mitigated attack variants. The proposed approach combines node location and trust-based parent selection. The result demonstrates the viability of the suggested countermeasure. In addition, demonstrating the superiority of the suggested countermeasure involves considering the precision of detecting the attack and the delay in isolating the attacker.

Agriculture is the main source of food. With thepassing of time, there are dangers in order to preserve on thefreshwater in agriculture sector. Thus, one of solutions to savethe freshwater is enhancing the wastewater. Machine learning(ML) algorithms are used in several applications, such as smartirrigation, to reduce freshwater loss via building highperformance ML algorithms. This paper proposes fouralgorithms: support vector machine (SVM), decision tree (DT),SVM with Adaboost, and DT with Adaboost to classify waterlevels of sprinklers for smart irrigation. Here, five levels ofwater are classified– Max, High, Medium, Low, and Stop. Theproposed algorithms are tested to obtain which algorithmachieves better performance and higher accuracy. Five stepssequentially are implemented on the used dataset via Pandasand Scikit-learn frameworks. The steps are preprocessing data,feature selection, feature scaling, training, and classification; toanalyze the performance of the algorithms. The results showedthat the DT algorithm with Adaboost is the best algorithmcompared to the rest of the algorithms. The DT algorithmachieves an accuracy score of 0.912 with a shorter testing timeof 2.2 seconds and mean square error (MSE) of 0.08.

Machine learning and deep learning algorithms have become increasingly important in the medical field, especially for diagnosing disease using medical databases. Techniques developed within these two fields are now used to classify different diseases. Although the number of Machine Learning algorithms is vast and increasing, the number of frameworks and libraries that implement them is also vast and growing. TensorFlow is a well-known machine learning library that has been used by several researchers in the field of disease classification. With the help of TensorFlow (Google's framework), a complex calculation can be addressed effectively by modeling it as a graph and properly mapping the graph segments to the machine in the form of a cluster. In this review paper, the role of the TensorFlow-Python framework- for disease classification is discussed.

Gastrointestinal (GI) system diseases have increased significantly, where colon and rectum cancer is considered the second cause of death in 2020. Wireless Capsule Endoscopy (WCE) is a revolutionary procedure for detecting Colorectal lesions. It was automatically used to detect the polyps, multiple SB lesions, bleeding, and Ulcer. The acquired video by the WCE can be processed using a Computer-Aided Diagnosis (CAD) system. However, such videos suffer several problems, including burling, high illumination. and distortion. These effects obligate the development of image processing techniques of high accuracy in detection using deep learning-based segmentation. In this paper, a transfer learning-based U-Net was proposed to transfer the knowledge between the medical images in the training phase and the subsequent segmentation using transfer learning to achieve better results and high accuracy results compared to other related studies. The improvement is done by using an algorism written in python code The results showed average segmentation accuracy of 98.67%

Copyright of Financial Research Journal (FRJ) is the property of University of Tehran, Faculty of Management and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)