Anomaly Detection Using Supervised learning Techniques in Social Networks DOI Creative Commons

Prof .Dr. Chezalina Binti Zulkifli

Wasit Journal of Computer and Mathematics Science, Год журнала: 2022, Номер 1(3), С. 16 - 20

Опубликована: Сен. 30, 2022

Intrusion detection corresponds to a suite of techniques that are used identify attacks against computers and network infrastructures. As the cost information processing Internet accessibility falls, more organizations becoming vulnerable wide variety cyber threats. Web mining based intrusion generally fall into one two categories; misuse anomaly detection. In detection, each instance in data set is labelled as ‘normal’ or ‘intrusive’ learning algorithm trained over data. These able automatically retrain models on different input include new types attacks, long they have been appropriately. Evaluation results show proposed approach can reduce number alerts by 94.32%, effectively improving alert management process. Because use ensemble optimal algorithms approach, it inform security specialist state monitored an online manner.

Язык: Английский

Gender Recognition of Human from Face Images Using Multi-Class Support Vector Machine (SVM) Classifiers DOI Open Access
Mohammed Jawad Al Dujaili, Haider TH. Salim ALRikabi, Nisreen Khalil Abed

и другие.

International Journal of Interactive Mobile Technologies (iJIM), Год журнала: 2023, Номер 17(08), С. 113 - 134

Опубликована: Апрель 26, 2023

In the realm of robotics and interactive systems, gender recognition is a crucial problem. Considering several uses it has in security, web search, human-computer interactions, etc., from facial photos garnered lot attention. The need to use enhance techniques felt more strongly today due significant development design systems. Relatively speaking other approaches, progress gained this area thus far not exceptional. Thus, novel method been adopted study improve accuracy comparison earlier research. To create best rate efficiency suggested research, we choose minimal set characteristics. Testing on FERET UTK-Face datasets reveals that our algorithm lower degree inaccuracy. article, input image person's face pre-processed extract right features once recognized. Gender separation achieved using Multi-class Support Vector Machine (SVM) Classifiers after normalized images have extracted Histogram Oriented Gradient (HOG), Gabor Filters, Speeded Up Robust Features (SURF), as well their combination select most appropriate feature them for classification. As reduction feature, Principal Component Analysis (PCA) also employed. Using proposed approach, 98.75% precision accomplished database runtime performance 0.4 Sec. database, 97.43% 0.5

Язык: Английский

Процитировано

19
Impact of Deep Learning Strategy in Mathematics Achievement and Practical Intelligence among High School Students DOI Open Access

Sabah Saeed Hammadi,

Ban Hassan Majeed, Areej Khuder Hassan

и другие.

International Journal of Emerging Technologies in Learning (iJET), Год журнала: 2023, Номер 18(06), С. 42 - 52

Опубликована: Март 21, 2023

— To identify the effect of deep learning strategy on mathematics achievement and practical intelligence among secondary school students during 2022/2023 academic year. In research, experimental research method with two groups (experimental control) a post-test were adopted. The community is represented by female fifth scientific grade from first Karkh Education Directorate. (61) intentionally chosen, they divided into groups: an group (30) who taught according to proposed strategy, control (31) usual method. For purpose collecting data for experiment, test was built, which in its final form (25) items out (20) objective type both them. Based findings, studied strategies outperformed those traditional.

Язык: Английский

Процитировано

13
Direction-Guided Sparse Representation Method for Finger Vein Recognition DOI
Lizhen Zhou, Lu Yang, Qinggang Meng

и другие.

Lecture notes in computer science, Год журнала: 2025, Номер unknown, С. 36 - 45

Опубликована: Янв. 1, 2025

Язык: Английский

Процитировано

0
Review Vehicular Ad hoc Networks Security Challenges and Future Technology DOI Creative Commons

Haifaa Ahmed Hassan,

Sahar Wahab khadim

Wasit Journal of Computer and Mathematics Science, Год журнала: 2022, Номер 1(3), С. 1 - 9

Опубликована: Сен. 30, 2022

Vehicular Ad hok Networks (VANET) is an emerging technology with both a bright future and significant concerns, particularly in terms of security. This study focuses on three-part VANET security frameworks. The first gives thorough review the needs, difficulties, characteristics In order to create secure architecture effective party communication, certain needs should be taken into account. We provide information current designs as well widely used standard protocols. second concentrates brand-new categorization various assaults described literature remedies that go them. third compares few these options using established standards for VANET. Then, assist researchers usage, we call attention many outstanding topics technological obstacles linked

Язык: Английский

Процитировано

20
Automatic Segmentation of Cervical Cells Based on Star-Convex Polygons in Pap Smear Images DOI Creative Commons
Yanli Zhao, Chong Fu, Wenchao Zhang

и другие.

Bioengineering, Год журнала: 2022, Номер 10(1), С. 47 - 47

Опубликована: Дек. 30, 2022

Cervical cancer is one of the most common cancers that threaten women's lives, and its early screening great significance for prevention treatment cervical diseases. Pathologically, accurate segmentation cells plays a crucial role in diagnosis cancer. However, frequent presence adherent or overlapping Pap smear images makes separating them individually difficult task. Currently, there are few studies on cells, existing methods commonly suffer from low accuracy complex design processes. To address above problems, we propose novel star-convex polygon-based convolutional neural network with an encoder-decoder structure, called SPCNet. The model accomplishes relying three steps: automatic feature extraction, polygon detection, non-maximal suppression (NMS). Concretely, new residual-based attentional embedding (RAE) block suggested image extraction. It fuses deep features attention-based layers shallow original through residual connection, enhancing network's ability to extract abundant features. And then, adaptive NMS (PA-NMS) algorithm adopted screen generated proposals further achieve detection thus allowing completely segment cell instances images. Finally, effectiveness our method evaluated independent datasets. Extensive experimental results demonstrate obtains superior performance compared other well-established algorithms.

Язык: Английский

Процитировано

16
The Effect of Cognitive Modeling in Mathematics Achievement and Creative Intelligence for High School Students DOI Open Access
Ban Hassan Majeed, Areej Khuder Hassan,

Sabah Saeed Hammadi

и другие.

International Journal of Emerging Technologies in Learning (iJET), Год журнала: 2023, Номер 18(09), С. 203 - 215

Опубликована: Май 10, 2023

To identify the impact of cognitive modeling in mathematics achievement, creative intelligence for secondary school students during 2022/2023 academic year. achieve aim research, researchers adopted experimental research method with two groups (experimental; control) a post-test. Community was identified, which represented by second intermediate grade /first KED. (70) male were intentionally chosen as sample study, and they divided into groups: an who taught according to modeling, other control group studied usual method. Equivalence made between chronological age, level intelligence, previous achievement mathematics. For purpose collecting data experiment, test built its final form (20) also out (25) paragraph from thematic type both them. Appropriate statistical analyzes carried out, besides done making sure psychometric properties tests. Based on findings, excelled over those traditional.

Язык: Английский

Процитировано

5
A Review of Hash Function Types and their Applications DOI Creative Commons

Haifaa Ahmed Hasan,

Hassan F. Al-Layla,

Farah N. Ibraheem

и другие.

Wasit Journal of Computer and Mathematics Science, Год журнала: 2022, Номер 1(3), С. 75 - 88

Опубликована: Сен. 30, 2022

In the recent decade, global community has been one of world's most prominent targets terrorist strikes. There is still study being done and improved efficiency procedures needed in this sector, however security forces law enforcement agencies are creating launching high weaponry to secure their territories. a huge necessity include greater level protection into military vehicles wake several attacks sniffing officers. Terrorists rely heavily on infiltration while avoiding detection by measures because holes these hypothetical situations. This paper presents central methods hash functions, cryptography dynamic encryption that may be utilised personnel increase safety, privacy, resistance communications with another. article details many algorithms included laser guided defensive weapons provide safe communication across system.

Язык: Английский

Процитировано

8
An optimized deep learning based depthwise separable MobileNetV3 approach for automatic finger vein recognition system DOI

Sambhaji Vamanrao Deshmukh,

Nitish Zulpe

Multimedia Tools and Applications, Год журнала: 2024, Номер 83(24), С. 64285 - 64313

Опубликована: Янв. 15, 2024

Язык: Английский

Процитировано

1
EnhanceDeepIris Model for Iris Recognition Applications DOI Creative Commons
Shouwu He, Xiaoying Li

IEEE Access, Год журнала: 2024, Номер 12, С. 66809 - 66821

Опубликована: Янв. 1, 2024

In this study, an iris recognition technique based on enhanced EnhanceDeepIris model is proposed in order to examine at a deeper level. The process first uses convolutional features extract human features. Then, solve the deformation problem caused by radial data of texture, sequential metric introduced realize effective iris. Finally, structure thoroughly examined investigate its actual performance. study found that after running four algorithms ND-IRIS-0405 and CASIA-Lamp datasets, loss function value research method began approach lowest 14th 9th iterations, while other continued slowly decrease. Additionally, 5th 4th iterations respectively, accuracy was nearly 91.00%.After 20 classification predictions, average accuracies paper, segmentation end-to-end multi-task network IrisST-Net, full complex-valued neural network, with hybrid preprocessing feature extraction were 99.88%, 98.72%, 97.47%, 89.77%, respectively. These findings suggest recognizes best can identify primary contour information accuracy. results application demonstrate provides clearest detection edge, less noise, accurately detect main This reference for optimizing related technologies field image recognition.

Язык: Английский

Процитировано

1
Finger Vein Extraction and Authentication for Security Purpose Based on CNN DOI
Rajan Nathan,

S Sakthivel,

P. Selvi Rajendran

и другие.

Опубликована: Апрель 18, 2024

Finger Vein ID is a biometric verification toolthat compares person's finger vascular pattern to previously gathered data. Numerous applications, such as automated teller machines, computer and network authentication, end point security, fingerprint recognition-one of the numerous biometrics used identify people-are now being developed or deployed for this technology. Individuals validate their identities that uses images vein patterns in human fingers lie beneath skin's surface employeetime attendance tracking, credit card vehicle security. With advancement electronics, there has been steady rise need easy-to-use, convenient, highly secure authentication solutions safeguard personal data held mobile devices. One most effective identification methods user recognition technique. due its security singularity. Since dawn technology, created detection veins; however, majority these techniques relied on manually constructed characteristics, which have limitations with regard image quality, location, etc. neural networks initially appeared, identity created. This research identifies variables may impact outcome surveys many utilizing CNN. The experimental examination both user- designed datasets different qualities publicly available from veins. Biometrics, method protecting information, employs physiological behavioral traits humans uniquely them. We send over SMS using AVR technology deliver verified individual.

Язык: Английский

Процитировано

1