Solving Engineering Optimization Problems Based on Multi-Strategy Particle Swarm Optimization Hybrid Dandelion Optimization Algorithm

Biomimetics, Journal Year: 2024, Volume and Issue: 9(5), P. 298 - 298

Published: May 17, 2024

In recent years, swarm intelligence optimization methods have been increasingly applied in many fields such as mechanical design, microgrid scheduling, drone technology, neural network training, and multi-objective optimization. this paper, a multi-strategy particle hybrid dandelion algorithm (PSODO) is proposed, which based on the problems of slow speed being easily susceptible to falling into local extremum ability algorithm. This makes whole more diverse by introducing strong global search unique individual update rules (i.e., rising, landing). The ascending descending stages also help introduce changes explorations space, thus better balancing search. experimental results show that compared with other algorithms, proposed PSODO greatly improves optimal value ability, convergence speed. effectiveness feasibility are verified solving 22 benchmark functions three engineering design different complexities CEC 2005 comparing it algorithms.

Language: Английский

---

Enhancing skin lesion classification: a CNN approach with human baseline comparison

PeerJ Computer Science, Journal Year: 2025, Volume and Issue: 11, P. e2795 - e2795

Published: April 15, 2025

This study presents an augmented hybrid approach for improving the diagnosis of malignant skin lesions by combining convolutional neural network (CNN) predictions with selective human interventions based on prediction confidence. The algorithm retains high-confidence CNN while replacing low-confidence outputs expert assessments to enhance diagnostic accuracy. A model utilizing EfficientNetB3 backbone is trained datasets from ISIC-2019 and ISIC-2020 SIIM-ISIC melanoma classification challenges evaluated a 150-image test set. model’s are compared against 69 experienced medical professionals. Performance assessed using receiver operating characteristic (ROC) curves area under curve (AUC) metrics, alongside analysis resource costs. baseline achieves AUC 0.822, slightly below performance experts. However, improves true positive rate 0.782 reduces false 0.182, delivering better minimal involvement. offers scalable, resource-efficient solution address variability in image analysis, effectively harnessing complementary strengths humans CNNs.

Language: Английский

---

Derin Öğrenme ve Özellik Seçimi Yaklaşımları Kullanılarak Göz Hastalıkları Tespiti

DÜMF Mühendislik Dergisi, Journal Year: 2024, Volume and Issue: unknown

Published: May 1, 2024

Göz sağlığı, önemli bir halk sağlığı konusudur ve göz hastalıkları dünya çapında ciddi sağlık sorunlarına neden olmaktadır. hastalıkları, görme yeteneğini etkileyebilen yaşam kalitesini ölçüde azaltabilen çeşitli sorunlarıdır. Bunlar arasında normal glukom, diyabetik retinopati katarakt yer tutmaktadır. Bu hastalıkların erken tanınması uygun tedavi yöntemlerinin uygulanması, sağlığının korunması kayıplarının en aza indirilmesi açısından hayati öneme sahiptir. Son dönemlerde, hastalıklarının teşhisi için yapay zekâ tekniklerinin kullanımı yaygınlaşmaktadır. teknikler, görüntü analizi derin öğrenme gibi ileri algoritmaları içerir tedavisi araç haline gelmektedir. çalışmada, fundus görüntülerinden doğru teşhis edilmesi özellik seçimi kombinasyonu yoluyla metasezgisel yöntemlerle optimize edilmiş metodoloji geliştirilmiştir. dört sınıflı veri setinden elde edilen görüntüler üzerinde çıkarımı önceden eğitilmiş mimarileri olan ResNet101, DenseNet201 DarkNet53 kullanılmıştır. mimarilerden özellikler birleştirilerek hibrit havuzu oluşturulmuştur. Oluşturulan bu havuz, görüntülerin daha etkili şekilde temsil edilmesini sağlamak Elde özelliklerin içinden önemsiz olanları elemek optimizasyon yöntemi parçacık sürü optimizasyonu (PSO) Görüntülerin sınıflandırılması makine öğrenmesi yöntemlerinden destek vektör makinesi (SVM) tercih edilmiştir. SVM'in performansını artırmak amacıyla, hiperparametrelerin Bayesian tekniği, SVM'nin iyi ayarlanmasına setine uyum sağlamasına yardımcı olmuştur. Deneysel çalışmaların sonuçlarına göre, sınıflandırma doğruluğu %93.8 olarak belirlenmiştir. sonuçlar, önerilen yöntemin tespitinde kullanılabileceğini göstermektedir. çalışma, zeka tıbbi görüntüleme alanında rol oynayabileceğini teşhisinde kullanılabilecek potansiyel olduğunu vurgulamaktadır.

---

Integrated bagging-RF learning model for diabetes diagnosis in middle-aged and elderly population

PeerJ Computer Science, Journal Year: 2024, Volume and Issue: 10, P. e2436 - e2436

Published: Oct. 31, 2024

As the population ages, increase in number of middle-aged and older adults with diabetes poses new challenges to allocation resources healthcare system. Developing accurate prediction models is a critical public health strategy improve efficient use ensure timely effective treatment. In order identification patients, Bagging-RF model proposed. study, two datasets on Kaggle were first preprocessed, including unique heat coding, outlier removal, age screening, after which data categorized into three groups, 50–60, 60–70, 70–80, balanced using SMOTE technique. Then, machine learning classifiers trained integrated eight other classifiers. Finally, model’s performance was evaluated by accuracy, F 1 score, metrics. The results showed that outperformed classifiers, exhibiting 97.35%, 95.55%, 95.14% accuracy Score at Diabetes Prediction Dataset for groups 70–80; 97.03%, 94.90%, 93.70% Dataset. 95.13% Score; accuracy; 94.89%, addition, while models, such as ET, RF, Adaboost, XGB, fail outperform Bagging-RF, they also show excellent performance.

Language: Английский

---

Automatic Detection of Gastrointestinal Diseases Using Wireless Capsule Endoscopy Images

NATURENGS MTU Journal of Engineering and Natural Sciences Malatya Turgut Ozal University, Journal Year: 2024, Volume and Issue: unknown

Published: Nov. 26, 2024

Gastrointestinal (GI) diseases are various disorders related to the digestive system. This system includes esophagus, stomach, small and large intestines, liver, gallbladder pancreas, starting from mouth. Early diagnosis is very important in treatment of disease. The earlier disease diagnosed, higher chance patient being treated. In recent years, it known that artificial intelligence techniques have been widely used classification. Among techniques, deep learning methods produce successful results image classification frequently used. success has tried be GI diseases. Within scope this study, was detect bleeding or lesions publicly available wireless capsule endoscopy (WCE) images. As a result experiments, 5 different architectures were Features extracted two showed highest accuracy combined. Neighborhood Component Analysis (NCA) dimension reduction method applied obtained feature map hybrid model obtained. It seen proposed achieved an value 86.3%.

Language: Английский

---