International Journal of Computational Learning & Intelligence

Enhancing Digital Circuit Performance Using Memristor-Inspired Amplifiers

Thamatam Venkata Chalama Reddy — 2025-04-02

In this paper, we present the design, implementation, and evaluation of digital logic circuits using Memristor-based technology. The focus is on basic gates, a 2 × 1 multiplexer (MUX), a full adder, a full subtractor, and an amplifier, all implemented using the Cadence Virtuoso platform. The Memristor model employed here shows significant improvements in power efficiency, area reduction, and speed compared to traditional 45-nm CMOS technologies. Our results demonstrate that Memristor-based circuits can achieve up to 71.4% reduction in area, 40% reduction in power consumption, and 54% reduction in delay, highlighting the potential of Memristor technology for future low-power, high-performance digital systems.

A Comprehensive Multi-Modal Framework For Cyberbullying Detection On Social Media

M Suchithra — 2025-04-02

The pervasive use of social media has led to an alarming rise in cyberbullying, particularly among younger users, posing significant threats to mental and emotional well-being. Traditional approaches to cyberbullying detection have predominantly focused on textual analysis, which often fails to capture the multi-modal nature of bullying content, including images, videos, and contextual metadata. To address this limitation, we propose a novel multi-modal cyberbullying detection framework that integrates textual, visual, and contextual information to identify bullying behavior more effectively. Our approach leverages advanced deep learning techniques, including Hierarchical Attention Networks (HAN) and Bidirectional Long Short-Term Memory (BiLSTM) networks, to model the complex interactions between different modalities. The framework processes user-generated posts, combining text and image data, along with metadata such as timestamps and user interactions, to predict whether a post constitutes cyberbullying. This research provides a robust, scalable solution for identifying and mitigating harmful content on social networks

Smart Water Leakage Detection And Prevention System Using IoT Technology

G Sateesh — 2025-04-03

The "Water Leakage Detection System Using IoT" is designed to detect and prevent water leakage by utilizing a Raspberry Pi Model B+ as the central control unit. It integrates several sensors, including a soil moisture sensor to assess soil dampness, a DHT11 sensor to measure temperature and humidity, and a flow sensor to monitor water flow. When any of these parameters deviate from the expected range such as an increase in temperature, abnormal moisture levels, or irregular water flow the system triggers an alert mechanism.A GSM module is employed to send an SMS notification to the user, ensuring they are promptly informed of any anomalies. Simultaneously, a buzzer sounds to draw immediate attention to the issue. An LCD screen provides real-time data and system status, while a 5mm red LED lights up when a critical parameter exceeds its threshold, indicating a potential leakage or abnormal condition. The system is powered by a reliable 12V 1A adapter, ensuring continuous and stable operation, making it suitable for real-time monitoring and response. Furthermore, the system utilizes machine learning algorithms to analyze sensor data and improve detection accuracy over time. By learning from historical data, it enhances its ability to identify patterns associated with leaks or irregular conditions. This proactive approach helps minimize water waste and prevent damage, making the system an efficient and intelligent solution for water leakage detection in various environments.

Age and Gender Prediction Using Swin Transformer and Multitasking Learning

B Sailendra Reddy — 2025-04-03

Age and gender prediction from facial images is an essential task in applications such as security systems human-computer interaction and personalized recommendations however variations in facial features due to lighting expressions and aging effects make it a challenging problem traditional convolutional neural networks CNNs often struggle with generalization whereas transformer-based models have shown superior performance by capturing long-range dependencies through self-attention mechanisms a multi-task learning approach where age estimation gender classification and contextual age positioning are trained together enhances feature representation and improves accuracy incorporating feature reweighting techniques allows the model to focus on critical facial attributes refining predictions dynamically additionally leveraging contextual learning such as relative age positioning strengthens the models ability to understand relationships between different age groups evaluations using benchmark datasets with diverse demographic distributions demonstrate the effectiveness of such an approach with performance measured through metrics like mean absolute error MAE for age estimation and classification accuracy for gender prediction future research can further enhance these models by integrating domain adaptation techniques and optimizing computational efficiency for real-time applications in biometric authentication healthcare and social media analytics

Analysing Mental Health Through Social Media and Computational Linguistics

Shaik Afreen — 2025-04-03

Mental health disorders are a growing concern in today’s digital age, with social media platforms serving as a reflection of user’s mental health states.

It is crucial to explore the underlying causes that drive individuals of all ages toward depression and identify effective ways to encourage them to choose life. In today's digital era, social media serves as a significant platform where people express their emotions, daily activities, and thoughts. This has led to the question of whether analyzing social media content can help determine an individual's emotional state, particularly identifying distress levels that may indicate suicidal tendencies.

This research explores the application of computational linguistics and machine learning techniques to predict mental health conditions based on social media text input. The system processes user-generated content and timestamps to classify mental health states using various classifiers, including Convolutional Neural Networks (CNN), Support Vector Machines (SVM), Logistic Regression, and others. Our approach leverages natural language processing (NLP) and deep learning models to analyze linguistic patterns associated with mental health indicators such as depression, anxiety, and stress. The proposed framework offers a novel, automated method for early mental health assessment, contributing to digital mental health monitoring and intervention strategies.