Preferred Language
Articles
/
fBb2j4oBVTCNdQwCuJ_G
Deep Transfer Learning for Improved Detection of Keratoconus using Corneal Topographic Maps
...Show More Authors
Abstract <p>Clinical keratoconus (KCN) detection is a challenging and time-consuming task. In the diagnosis process, ophthalmologists must revise demographic and clinical ophthalmic examinations. The latter include slit-lamb, corneal topographic maps, and Pentacam indices (PI). We propose an Ensemble of Deep Transfer Learning (EDTL) based on corneal topographic maps. We consider four pretrained networks, SqueezeNet (SqN), AlexNet (AN), ShuffleNet (SfN), and MobileNet-v2 (MN), and fine-tune them on a dataset of KCN and normal cases, each including four topographic maps. We also consider a PI classifier. Then, our EDTL method combines the output probabilities of each of the five classifiers to obtain a decision based on the fusion of probabilities. Individually, the classifier based on PI achieved 93.1% accuracy, whereas the deep classifiers reached classification accuracies over 90% only in isolated cases. Overall, the average accuracy of the deep networks over the four corneal maps ranged from 86% (SfN) to 89.9% (AN). The classifier ensemble increased the accuracy of the deep classifiers based on corneal maps to values ranging (92.2% to 93.1%) for SqN and (93.1% to 94.8%) for AN. Including in the ensemble-specific combinations of corneal maps’ classifiers and PI increased the accuracy to 98.3%. Moreover, visualization of first learner filters in the networks and Grad-CAMs confirmed that the networks had learned relevant clinical features. This study shows the potential of creating ensembles of deep classifiers fine-tuned with a transfer learning strategy as it resulted in an improved accuracy while showing learnable filters and Grad-CAMs that agree with clinical knowledge. This is a step further towards the potential clinical deployment of an improved computer-assisted diagnosis system for KCN detection to help ophthalmologists to confirm the clinical decision and to perform fast and accurate KCN treatment.</p>
Scopus Clarivate Crossref
View Publication
Publication Date
Wed Jan 01 2025
Journal Name
Iv. International Rimar Congress Of Pure, Applied Sciences
A New Intrusion Detection Approach Based on RNA Encoding and K-Means Clustering Algorithm Using KDD-Cup99 Dataset
...Show More Authors

Intrusion detection systems (IDS) are useful tools that help security administrators in the developing task to secure the network and alert in any possible harmful event. IDS can be classified either as misuse or anomaly, depending on the detection methodology. Where Misuse IDS can recognize the known attack based on their signatures, the main disadvantage of these systems is that they cannot detect new attacks. At the same time, the anomaly IDS depends on normal behaviour, where the main advantage of this system is its ability to discover new attacks. On the other hand, the main drawback of anomaly IDS is high false alarm rate results. Therefore, a hybrid IDS is a combination of misuse and anomaly and acts as a solution to overcome the dis

... Show More
Preview PDF
Publication Date
Mon Feb 01 2016
Journal Name
Swarm And Evolutionary Computation
Improving the performance of evolutionary multi-objective co-clustering models for community detection in complex social networks
...Show More Authors

Scopus (34)
Crossref (29)
Scopus Clarivate Crossref
Publication Date
Sat Apr 01 2023
Journal Name
International Journal Of Electrical And Computer Engineering
Intrusion detection method for internet of things based on the spiking neural network and decision tree method
...Show More Authors

The prevalence of using the applications for the internet of things (IoT) in many human life fields such as economy, social life, and healthcare made IoT devices targets for many cyber-attacks. Besides, the resource limitation of IoT devices such as tiny battery power, small storage capacity, and low calculation speed made its security a big challenge for the researchers. Therefore, in this study, a new technique is proposed called intrusion detection system based on spike neural network and decision tree (IDS-SNNDT). In this method, the DT is used to select the optimal samples that will be hired as input to the SNN, while SNN utilized the non-leaky integrate neurons fire (NLIF) model in order to reduce latency and minimize devices

... Show More
Preview PDF
Publication Date
Thu Mar 30 2023
Journal Name
Iraqi Journal Of Science
A Tri-Gene Ontology Migration Operator for Improving the Performance of Meta-heuristics in Complex Detection Problems
...Show More Authors

      Detecting protein complexes in protein-protein interaction (PPI) networks is a challenging problem in computational biology. To uncover a PPI network into a complex structure, different meta-heuristic algorithms have been proposed in the literature. Unfortunately, many of such methods, including evolutionary algorithms (EAs), are based solely on the topological information of the network rather than on biological information. Despite the effectiveness of EAs over heuristic methods, more inherent biological properties of proteins are rarely investigated and exploited in these approaches. In this paper, we proposed an EA with a new mutation operator for complex detection problems. The proposed mutation operator is formulate

... Show More
Scopus (3)
Scopus Crossref
Publication Date
Sat Apr 01 2023
Journal Name
International Journal Of Electrical And Computer Engineering (ijece)
Intrusion detection method for internet of things based on the spiking neural network and decision tree method
...Show More Authors

The prevalence of using the applications for the internet of things (IoT) in many human life fields such as economy, social life, and healthcare made IoT devices targets for many cyber-attacks. Besides, the resource limitation of IoT devices such as tiny battery power, small storage capacity, and low calculation speed made its security a big challenge for the researchers. Therefore, in this study, a new technique is proposed called intrusion detection system based on spike neural network and decision tree (IDS-SNNDT). In this method, the DT is used to select the optimal samples that will be hired as input to the SNN, while SNN utilized the non-leaky integrate neurons fire (NLIF) model in order to reduce latency and minimize devices

... Show More
Scopus (19)
Crossref (9)
Scopus Crossref
Publication Date
Tue Apr 19 2022
Journal Name
Chemistryselect
A Continuous Flow Injection/Indirect Photometric Method for the Detection of Fosetyl Aluminum in Commercial Pesticide Formulations
...Show More Authors
Abstract<p>Because of their Physico‐chemical characteristics and its composition, the development of new specific analytical methodologies to determine some highly polar pesticides are required. The reported methods demand long analysis time, expensive instruments and prior extraction of pesticide for detection. The current work presents a new flow injection analysis method combined with indirect photometric detection for the determination of Fosetyl‐Aluminum (Fosetyl‐Al) in commercial formulations, with rapid and highly accurate determination involving only construction of manifold system combined with photometric detector without need some of the pre‐treatments to the sample before the analysis such a</p> ... Show More
View Publication
Scopus (6)
Crossref (3)
Scopus Clarivate Crossref
Publication Date
Sat Apr 09 2022
Journal Name
Engineering, Technology &amp; Applied Science Research
A Semi-Empirical Equation based on the Strut-and-Tie Model for the Shear Strength Prediction of Deep Beams with Multiple Large Web Openings
...Show More Authors

The behavior and shear strength of full-scale (T-section) reinforced concrete deep beams, designed according to the strut-and-tie approach of ACI Code-19 specifications, with various large web openings were investigated in this paper. A total of 7 deep beam specimens with identical shear span-to-depth ratios have been tested under mid-span concentrated load applied monotonically until beam failure. The main variables studied were the effects of width and depth of the web openings on deep beam performance. Experimental data results were calibrated with the strut-and-tie approach, adopted by ACI 318-19 code for the design of deep beams. The provided strut-and-tie design model in ACI 318-19 code provision was assessed and found to be u

... Show More
View Publication
Scopus (6)
Crossref (8)
Scopus Crossref
Publication Date
Sun Nov 09 2025
Journal Name
Journal Of Physical Education
The effect of the Perkins-Blyth model on learning some compound skills in soccer for second intermediate students
...Show More Authors

View Publication
Publication Date
Thu Apr 06 2023
Journal Name
International Journal Of Emerging Technologies In Learning (ijet)
The Impact of a Scenario-Based Learning Model in Mathematics Achievement and Mental Motivation for High School Students
...Show More Authors

Crossref (3)
Crossref
Publication Date
Mon Jun 01 2020
Journal Name
مجلة العلوم و التكنولوجية للنشاطات البدنية و الرياضية/ الجزائر
The effect of using the five-finger strategy on learning movement chain on a balance beam in the artistic gymnastics of women
...Show More Authors

Abstract: The aim of the research identify the effect of using the five-finger strategy in learning a movement chain on the balance beam apparatus for students in the third stage in the College of Physical Education and Sports Science, as well as to identify which groups (experimental and controlling) are better in learning the kinematic chain on the balance beam device, has been used The experimental approach is to design the experimental and control groups with pre-and post-test. The research sample was represented by third-graders, as the third division (j) was chosen by lot to represent the experimental group, and a division Third (i) to represent the control group, after which (10) students from each division were tested by lot to repr

... Show More
View Publication Preview PDF