Classifying cuneiform symbols using machine learning algorithms with unigram features on a balanced dataset

Journal of Intelligent Systems, Journal Year: 2023, Volume and Issue: 32(1)

Published: Jan. 1, 2023

Abstract Problem Recognizing written languages using symbols in cuneiform is a tough endeavor due to the lack of information and challenge process tokenization. The Cuneiform Language Identification (CLI) dataset attempts understand seven dialects, including Sumerian six dialects Akkadian language: Old Babylonian, Middle Babylonian Peripheral, Standard Neo-Babylonian, Late Neo-Assyrian. However, this suffers from problem imbalanced categories. Aim Therefore, article aims build system capable distinguishing between several solving unbalanced categories CLI dataset. Methods Oversampling technique was used balance dataset, performance machine learning algorithms such as Support Vector Machine (SVM), K-Nearest Neighbors (KNN), Decision Tree (DT), Random Forest (RF), deep neural networks (DNNs) unigram feature extraction method investigated. Results proposed (SVM, KNN, DT, RF) on balanced obtained an accuracy 88.15, 88.14, 94.13, 95.46%, respectively, while DNN model got 93%. This proves improved compared related works. Conclusion improvement classifiers when working use features also showed classifier it reduced size data accelerated processing process.

Language: Английский

---

Using an innovative method for breast cancer diagnosis based on Extreme Gradient Boost optimized by Simplified Memory Bounded A*

Biomedical Signal Processing and Control, Journal Year: 2023, Volume and Issue: 87, P. 105450 - 105450

Published: Oct. 1, 2023

Language: Английский

---

Two-and-a-half order score-based model for solving 3D ill-posed inverse problems

Computers in Biology and Medicine, Journal Year: 2023, Volume and Issue: 168, P. 107819 - 107819

Published: Dec. 5, 2023

Language: Английский

---

Ensemble of Deep Features for Breast Cancer Histopathological Image Classification

The Computer Journal, Journal Year: 2024, Volume and Issue: 67(6), P. 2126 - 2136

Published: Jan. 14, 2024

Abstract Analysis of histopathological images (HIs) is crucial for detecting breast cancer (BR). However, because they vary, it still very difficult to extract well-designed elements. Deep learning (DL) a recent development that used high-level features. DL techniques continue confront several problems, such as the need sufficient training data models, which reduces classification findings. In this study, an ensemble deep transfer convolutional neural network presented address problem. The pre-trained models (ResNet50 and MobileNet) are employed features by freezing front layer parameters while fine-tuning last layers. proposed framework, KNN, SVM, logistic regression networks base classifiers. majority vote product approaches integrate predictions each separate classifier. benchmark BreaKHis dataset, suggested model compared some current approaches. It demonstrates obtains considerable accuracy 97.72% multiclass test, achieves 99.2% binary task. model’s effectiveness in extracting useful BR demonstrated comparison with existing cutting-edge models.

Language: Английский

---

Deep Learning Techniques in the Cancer-Related Medical Domain: A Transfer Deep Learning Ensemble Model for Lung Cancer Prediction

Baghdad Science Journal, Journal Year: 2023, Volume and Issue: 21(3), P. 1101 - 1101

Published: Aug. 20, 2023

المشكلة: يعتبر السرطان أحد أكثر الأمراض فتكًا في العالم. يمكن أن يسهل التعلم الآلي وخوارزميات العميق طريقة التعامل مع لا سيما مجال الوقاية من واكتشافه. إن للطرق التقليدية لتحليل بيانات محددات عديدة كما أنً تنمو بسرعة. هذا يجعل الممكن للتعلم المضي قدمًا بقدراته القوية على تحليل ومعالجة السرطان. الأهداف: تم الدراسة الحالية تقديم نظام دعم طبي للتنبؤ بسرطان الرئة. الطرائق: تستخدم ثلاثة نماذج مختلفة (EfficientNetB3 وResNet50 وResNet101) مفهوم نقل التعلم. تدريب النماذج الثلاثة باستخدام مجموعة لسرطان الرئة بالأشعة المقطعية المؤلفة 1000 صورة وأربع فئات للسرطان. تطبيق عملية Data Augmentation لحل مشكلة Overfitting وزيادة حجم البيانات وتعزيز التدريب. أيضًا استخدام الدمج مستوى الدرجة والتعلم التجميعي Ensemble Learning للحصول أفضل أداء وحل الدقة المنخفضة. تقييم جميع معايير وPrecision وRecall وF1-score. النتائج: تظهر التجارب الأداء العالي لنموذج بدقة 99.44٪، وهو المنهجيات الحديثة الحالية. الخلاصة: نتائج العالية والمتانة لنقل المجموعة المقترحة تعلم التحويل المختلفة.

---