i-manager's Journal on Artificial Intelligence & Machine Learning (JAIM)


Volume 1 Issue 2 July - December 2023

Research Paper

Uncovering the Truth: A Deep Learning Approach to Detecting Fake News

Bishal Bose*
Department of Computer Science, University of Mumbai, Navi Mumbai, India.
Bose, B. (2023). Uncovering the Truth: A Deep Learning Approach to Detecting Fake News. i-manager’s Journal on Artificial Intelligence & Machine Learning, 1(2),1-11. https://doi.org/10.26634/jaim.1.2.19771

Abstract

Fake news is a growing problem on social media and can have significant negative consequences for individuals and society owing to its accessibility, low cost, and quick distribution. It is difficult to automatically identify bogus news that defies the current content-based analysis techniques. One of the key reasons is that current NLP algorithms still lack common sense, which is frequently necessary for understanding how to read the news. Recent research has demonstrated that the propagation patterns of true and fake news differ on social media. With the potential for automatic fake news detection, this study investigates the use of Deep Learning (DL) models to detect fake news on social media. A Neural Network (NN) model was developed using Natural Language Toolkit (NLTK), TensorFlow, and Natural Language Processing (NLP) for textual analysis. The model was trained on a dataset of fake and real news articles, and was able to achieve high accuracy in identifying fake news. The results of this study suggest that DL models can be valuable tools for detecting fake news on social media.

Research Paper

A Distinctive Ensemble Deep Learning Model for Brain Tumor MRI Image Classification

Narasimha Rao Thota* , D. Vasumathi**
* Department of Computer Science and Engineering, JNTUK University, Kakinada, Andhra Pradesh, India.
** Department of Computer Science and Engineering, JNTUH University, Kukatpally, Hyderabad, Telangana, India.
Thota, N. R., and Vasumathi, D. (2023). A Distinctive Ensemble Deep Learning Model for Brain Tumor MRI Image Classification. i-manager’s Journal on Artificial Intelligence & Machine Learning, 1(2), 12-21. https://doi.org/10.26634/jaim.1.2.19281

Abstract

Brain tumor detection is challenging for radiologists. The early detection of brain tumors is critical, and automated techniques are necessary to achieve this goal. In this study, an automated method is proposed to distinguish between malignant and non-cancerous brain Magnetic Resonance Images (MRI). An ensemble technique was proposed that includes two Deep Learning (DL) models, a 6-Convolutional Neural Network (6-CNN) model with an efficient end-to-end, and a pre-trained Residual Network50 (ResNet50) model. The MRI image classification was experimented in two directions: one using the average probabilities of the ensemble model, and the other considering the optimal weights of the ensemble model for a Support Vector Machine (SVM) classifier with different kernels. Two datasets, Harvard and Retrospective Image Registration Evaluation (RIDER), were used to evaluate the performance of the proposed model. The 6-CNN-ResNet-SVM model achieved the highest accuracy of 97.32% for 10-fold validation, with the remaining performance metrics being an Area Under the Curve (AUC) of 0.98%, sensitivity of 93.62%, specificity of 98%, False Negative Rate (FNR) of 0.06, and False Positive Rate (FPR) of 0.00 for the linear kernel. The proposed model can identify tumors more accurately and quickly than existing approaches.

Research Paper

Fire and Smoke Detection using YOLOv8

Vinay Kumar Jain* , Chitrangad Jain**
*-** Shri Shankaracharya Technical Campus, Bhilai, India.
Jain, V. K., and Jain, C. (2023). Fire and Smoke Detection using YOLOv8. i-manager’s Journal on Artificial Intelligence & Machine Learning, 1(2), 22-29. https://doi.org/10.26634/jaim.1.2.19849

Abstract

In smart cities, fire can have disastrous effects, destroying property and putting residents' lives in danger, making it difficult to identify fire in real time because of the accuracy and speed constraints of traditional fire detection techniques. To address this issue, an accurate and cost-effective system that can be used in almost any fire detection scenario was developed. A CNN was used to analyze live video from a fire monitoring system to identify fire. An object identification model for deep learning called You Only Look Once (YOLOv8) was used to detect fire. To identify and alert videos from CCTV footage, a dataset of video frames with flames is used. After pre-processing the data, CNN is used to build a Machine Learning (ML) model. The methodology adopted in this study demonstrated the ability to adjust to various situations.

Research Paper

A Novel Approach for Lung Cancer Detection using Deep Belief Networks

Sangeeta Devi* , Pranjal Maurya**, Rajan Kumar Yadav***, Munish Saran****, Upendra Nath Tripathi*****
*-***** Department of Computer Science, Deen Dayal Upadhyaya Gorakhpur University, Gorakhpur, Uttar Pradesh, India.
Devi, S., Maurya, P., Yadav, R. K., Saran, M., and Tripathi, U. N. (2023). A Novel Approach for Lung Cancer Detection using Deep Belief Networks. i-manager’s Journal on Artificial Intelligence & Machine Learning, 1(2), 30-36. https://doi.org/10.26634/jaim.1.2.20010

Abstract

Due to the increased awareness of lung cancer, researchers have created many algorithms that can recognise the disease in its early stages using a variety of Machine Learning (ML) techniques. Clinicians can manage incidental or screen-found ambiguous pulmonary nodules with the help of machine learning-based models for lung cancer prediction. Such methods might be able to lower the variability in nodule classification, enhance decision making, and eventually decrease the proportion of benign nodules that do not need to be followed. This study proposes a novel lung cancer detection method based on Magnetic Resonance Imaging (MRI). Using ML to classify features in MRI scans, this technology is useful for the early detection of lung cancer. The performance was further enhanced using featureselection methodologies. The images were divided into segments using the FBSO feature selection method, and deep learning techniques were used to analyze the three standard datasets, S1, S2 and S3. In this investigation, 98.9% classifier optimality and 96.7% accuracy were attained. This new approach demonstrated excellent dependability and was found to be the most effective classifier system compared with previous studies.

Review Paper

Unlocking Clinical Insights from Medical Images using Deep Learning

Ushaa Eswaran* , Vishal Eswaran**
* Department of Electronics and Communication Engineering, Indira Institute of Technology and Sciences, Markapur, Andhra Pradesh, India.
** Consumer Value Store Health Centre, Dallas, Texas, United States.
Eswaran, U., and Eswaran, V. (2023). Unlocking Clinical Insights from Medical Images using Deep Learning. i-manager’s Journal on Artificial Intelligence & Machine Learning, 1(2), 37-48. https://doi.org/10.26634/jaim.1.2.20044

Abstract

Medical imaging is fundamental to modern precision medicine and the analysis of complex image data requires sophisticated techniques. This review focuses on recent deep-learning techniques for medical image analysis across existing modalities such as X-rays, MRI, CT and ultrasound. The aim of this study was to review the current state-of-the-art methods, highlight proven applications in precision diagnostics and prognosis, analyze key challenges, and identify promising future directions for this rapidly advancing field. Convolutional Neural Networks (CNNs) enable transformative improvements in tasks such as classification, segmentation and detection compared with prior approaches. A wide range of real-world applications across diagnostic, interventional, prognostic and pharmaceutical settings has been presented. The salient challenges concerning model interpretability, multimodal integration, algorithmic robustness, workflow integration, and responsible ethical deployment were discussed. This synergy between medical imaging and artificial intelligence continues to unlock abundant clinically relevant insights latent in images and transforms datadriven precision medicine for patient benefit.

Review Paper

Generative Adversarial Networks: Techniques, Area Covered, and Challenges

Choubey Aakanksha S.* , Gajbhiye Samta**, Tiwari Rajesh***
*-** Department of Computer Science & Engineering, Shri Shankaracharya Institute of Engineering and Technology, Bhilai, Chhattisgarh, India.
*** CMR Engineering College, Medchal, Hyderabad, Telangana, India.
Aakanksha, S. C., Samta, G., and Rajesh, T. (2023). Generative Adversarial Networks: Techniques, Area Covered, and Challenges. i-manager’s Journal on Artificial Intelligence & Machine Learning, 1(2), 49-55. https://doi.org/10.26634/jaim.1.2.20025

Abstract

Generative Adversarial Networks (GANs) have emerged as powerful frameworks for generating realistic and diverse data samples in various domains including computer vision, natural language processing and audio synthesis. In this study, a comprehensive overview of GANs, their key components, the training process, and the evolution of the field are presented. Different variations in GAN architectures and their applications across domains are discussed. Furthermore, the challenges and open research directions in GANs, their stability, mode collapse, evaluation metrics, and ethical considerations are analysed. The aim of this review is to provide a comprehensive understanding of GANs and their strengths, limitations, and potential future developments.