Browsing by Author "Ahammad, Mejbah"

Now showing 1 - 3 of 3

A Proficient Approach to Detect Osteosarcoma Through Deep Learning
(Daffodil International University, 2022-04-20) Ahammad, Mejbah; Abedin, Mohammad Joynul; Khan, Md. Asiqur Rahman; Alim, Md. Abdul; Rony, Mohammad Abu Tareq; Alam, K.M. Rashedul; Reza, D. S. A. Aashiqur; Uddin, Iktear
Osteosarcoma is a life-threatening bone cancer that usually attacks young adults and children, independent of age. It habitually starts in quick-growing bone areas close to the ends of the arm or leg bones, such as the distal femur, proximal tibia, and proximal humerus. However, it can still be revealed in any bone, including the pelvis, jaw, and shoulder. The starting and the preeminent conclusion of any cancer are to identify the tumor as before long as conceivable, and it's moreover pertinent for Osteosarcoma. Osteosarcoma has a few arrange in its life cycle. The need of categorizing cancer patients into tall or short risk categories has prompted several research organizations in the biomedical and bioinformatics fields to consider using Profound Learning (Deep Learning) methodologies. Fast.ai, a Deep Learning Framework for enhancing the efficiency and accu-racy of osteosarcoma tumor categorization into tumor classes, is presented in this study (tumor vs non-tumor). At the conclusion of the study, we found that employing neural networks may provide excellent precision and capability in osteosarcoma classification and model comparison.
Beyond Words: Unraveling Text Complexity with Novel Dataset and a Classifier Application
(IEEE, 2023-02-27) Islam, Mohammad Shariful; Rony, Mohammad Abu Tareq; Saha, Pritom; Ahammad, Mejbah; Alam, Shah Md Nazmul; Rahman, Md Saifur
Text classification is a fundamental aspect of Natural Language Processing (NLP). This research presents a novel human-annotated English sentence dataset categorized into four classes (simple, complex, compound, complex-compound) containing 22331 sentences and a sophisticated sentence classifier tool offering the capability to analyze and classify sentences within English text with particular relevance to literature writing. This study explores its performance using three distinct feature representation methods: Bag-of-Words (BoW), Term Frequency-Inverse Document Frequency (TF-IDF), and Word Embedding Features. The study involves the evaluation of four machine learning and two deep learning classifier models. BoW combined with Support Vector Classifier (SVC) and Logistic Regression (LR) demonstrated impressive accuracy rates, excelling in distinguishing sentence complexity. Word Embedding Features, specifically LSTM and RNN, offer a more profound semantic representation. LSTM stands out with the highest accuracy of 98.03% and balanced precision and recall, yielding an average F1-score of 97%. RNN, slightly less accurate at 97.75%, nevertheless exhibits competence in grasping sentence structure dependencies. It offers valuable insights for practical applications and contributes to the broader understanding of sentence structures and semantics.
Optimization of Liver RNA-Seq Analysis through HRNet: Bridging Computational Techniques and Biological Understanding
(2025-06-10) Ahammad, Mejbah; Zaman, Nourin; Tenis, Rishan Hasan; Al Mamun Sakib, Abdullah; Niger, Tasnim; Mama, Tithee Chakma
In the digital era, the rapid proliferation of RNA sequencing (RNA-Seq) has uncovered vast datasets, particularly in liver-specific gene expression, yet the complexity of such high-dimensional data often obscures actionable insights. This research focuses on optimizing the analysis of liver RNA-Seq data through the innovative application of the Human RNA Network (HRNet) architecture, renowned for its proficiency in handling complex image and data patterns. Employing a refined dataset from the All RNA-seq and ChIP-seq sample and signature search (ARCHS4) platform, this study evaluates the effectiveness of HRNet alongside traditional machine learning and emerging deep learning techniques. The introduction of enhanced preprocessing, feature selection, and dimensionality reduction methods tailored to the intricacies of liver RNA-Seq data marks a pivotal advance in computational biology. Rigorous testing against conventional models demonstrates the superior capability of HRNet, achieving remarkable precision in identifying and analyzing gene expressions critical to understanding liver functions and disorders. The results showcase not only HRNet’s robust performance, with accuracy rates exceeding 98.5% in complex gene expression patterns, but also its potential in transforming the landscape of genomic research, facilitating more precise diagnostic and therapeutic strategies. This study underscores the necessity of integrating advanced computational models to effectively decipher the voluminous data generated by RNA-Seq, thereby enhancing the accuracy and applicability of genomic analyses in medical research.