International Research Journal of Multidisciplinary Technovation

Integration of Model Free Kinetics and Machine Learning to Evaluate the Thermal Stability Behaviour of Nano Enhanced Composite Phase Change Materials

Pavan Kumar K — 2025-11-15

The phase change materials (PCMs) used in medium-temperature thermal energy storage can store and release a large amount of latent heat. However, their use in real-world applications is often limited by poor thermal conductivity and degradation stability during repeated heating cycles. Therefore, improving the thermal reliability and kinetic behaviour of PCMs is essential for advancing high-performance energy storage systems. This research assessed the thermal stability and degradation kinetics of pure D-mannitol and its graphene nanoplatelet (GNP)-reinforced composites. It introduced a new combination of model-free kinetic modelling and machine-learning-based predictions for evaluating the thermal reliability of nano-enhanced PCMs. Five different PCM compositions (containing 0, 0.25, 0.5, 0.75, and 1 wt% GNP) were prepared using ultrasonic-assisted dispersion. Non-isothermal thermogravimetric analysis (TGA) was performed at five different heating rates (10–25 °C·min⁻¹), and the activation energies (Eₐ) were determined using the Kissinger–Akahira–Sunose (KAS), Flynn–Wall–Ozawa (FWO), and Starink model-free methods. The Starink model yielded the lowest Eₐ for pure D-mannitol (61.99 kJ·mol⁻¹ at α ≈ 0.1), while the KAS and FWO models provided mean values of 137.62 kJ·mol⁻¹ and 141.48 kJ·mol⁻¹, respectively. Incorporating GNP improved thermal stability across all compositions, with Eₐ ranging from 123.67 to 149.08 kJ·mol⁻¹. The random forest regression model achieved the best predictive accuracy (R² = 0.99, RMSE = 0.0002), outperforming both linear and polynomial models. Overall, the addition of graphene nanoplatelets significantly enhances both chemical and thermal stability, confirming their role as effective nano-additives for PCM improvement and enabling the design of thermally stable “smart” energy storage systems for solar and sustainable applications. The thermal stability of Graphene Nanoplatelets (GNP)-enhanced D-Mannitol was determined through non-isothermal TGA analysis at heating rates between 10–25 °C min-1. The decomposition behaviour of the GNP-enhanced D-Man was characterized using Model-Free Kinetic Methods and Machine Learning Models to determine the accuracy of predicting the thermal stability of the D-Man using the most significant thermal parameters.

Exploring Second-Order Nonlinear Optical Performance in Oxyquinolinium 3-carboxypropanoate: A Combined Experimental and DFT Study

Thirumurugan R — 2025-11-18

The crystallization of oxyquinolinium 3-carboxypropanoate (OXSU) was successfully achieved using a well-optimized slow evaporation approach. Although earlier experimental investigations have discussed its structural, thermal, and nonlinear optical (NLO) characteristics, the present study is a detailed computational analysis to clarify the electronic basis of its NLO response. X-ray diffraction (XRD) studies established the non-centrosymmetric (NCS) P2₁ crystal structure of OXSU, fulfilling the fundamental symmetry condition for second-order NLO activity. Hirshfeld surface (HS) analysis indicated that the intermolecular O–H^…O/N–H^…O hydrogen-bonding interactions were particularly important for stabilizing the acentric packing of OXSU for SHG activity. Vibrational spectroscopy (FTIR and FT-Raman) also provided evidence of both functional groups and hydrogen bonding interactions. Moreover, Mulliken charge analysis revealed evidence of charge redistribution during the formation of an asymmetrical charge density. Frontier molecular orbital (FMO) energy level calculations indicated an energy gap (ΔE = 3.82 eV) that encouraged intramolecular charge transfer (ICT). The molecular electrostatic potential (MEP) indicated regions of nucleophilic and electrophilic directions suggesting asymmetrical electron mobility. The material exhibits a high first hyperpolarizability (β = 16.94 × 10^-30 esu), which is 33 times that of urea. The second harmonic generation (SHG) response is 30 % greater compared to potassium dihydrogen phosphate (KDP) and demonstrate phase-matchable behavior. This study documents the structure-property (SP) relationships in OXSU, correlating its acentric crystalline packing, hydrogen bonding interactions and charge transfer characteristics with NLO performance. These results establish that OXSU is a highly efficient, phase matchable organic NLO material with significant potential for optoelectronic applications.

A Unified Model for Estimation of Reference Evapotranspiration Using an Assembly of Ensemble Learners Coupled with Swarm Intelligence Optimizers

Gouravmoy Banerjee — 2025-10-14

Several machine learning models and their ensembles have been suggested for reference evapotranspiration (ET₀) modeling at different climatic regions. Researchers reported that optimizing model hyperparameters using an intelligent algorithm significantly improves the performance of such models. However, ensemble models hybridized with hyperparameter optimizers have hardly been applied for the precise estimation of ET₀ worldwide. The current research is devoted to designing sixteen hybrid versions of four ensemble models, alternatively coupled with four popular swarm intelligence optimization algorithms and finding the best-fit model against different input combinations of available climatic parameters for the groundwater-stressed region of North Bengal, India. The performances of four ensemble models and their sixteen hybrid versions were compared in terms of four well-recognized statistical metrics: the coefficient of determination (R²), Nash-Sutcliffe efficiency (NSE), root mean squared error (RMSE), and mean absolute error (MAE). Experimental results depicted that in nearly 92% of cases, the hybrid versions outperformed the primary ensemble models, irrespective of the available climatic parameters. In most cases, the ensemble models hybridized with the whale optimization algorithm (WOA) produced the highest estimation accuracy, followed by the sailfish optimizer (SFO). Solar radiation was also found to be the most significant climatic parameter for estimating ET₀ in this region.

YOLO-Based Generalized Framework for Leukemia Cell Detection Using Unified Microscopic Image Datasets

Ratnamala Mantri (Paswan) — 2025-10-15

Acute lymphoblastic leukemia is a kind of blood cancer that attacks the lymphoblast, a subgroup of white blood cells. Leukemia is a potentially lethal hematological cancer that requires prompt diagnosis. A skilled manual blood smear examination is one of the laborious and prone to human error conventional diagnostic methods. Although current automated methods developed by researchers use either single-cell or multi-cell pictures to detect leukemia cells, they frequently lack model generalization that perform better on heterogeneous datasets. They are also insufficient for deployment in real time. This study aims to develop generalized real-time system for detecting ALL cells from single and multi-cell microscopic blood smear images. The system utilizes three YOLO based state-of-the-art models: YOLO11, YOLOv8 and YOLOv5. The core novelty of this study lies in the creation of a unified dataset that integrates both single-cell and multi-cell microscopic blood smear images, this enables the model to learn generalized representations from diverse image contexts. Three datasets are merged to create the unified dataset, ALL-IDB1: multi- cell images, ALL-IDB2 & C-NMC-19: single-cell images. Image annotation and preprocessing are performed using Roboflow platform, while Google Colab is used for training and testing. These models are trained separately on individual datasets and the unified dataset.The performance of generalized YOLO models is assessed and contrasted against dataset-specific models using mAP@50 and recall metrics on the same set of unseen images from all three datasets.The experimental results indicate that generalized YOLOv8 model achieved notably high recall and competitive map@50, demonstrating strong adaptability and accuracy. These results highlight YOLOv8 as a promising solution for developing generalized model for leukemia cell detection.

Detection and Classification of Genetic Acute Myeloid Leukemia Cells using Deep Learning Techniques

Hema Patel — 2025-10-16

Acute Myeloid Leukemia (AML) is a hematological disease that is defined by the fast growth of aberrant myeloid precursor cells in the blood and bone marrow, which disrupts normal hematopoiesis. Treatment and prognosis are influenced by the early detection of this deadly illness and its appropriate classification. Therefore, utilizing the Human Leukemia Cytomorphology Collection dataset, which comprises leukemic and normal single-cell images of Acute Myeloid Leukemia (AML) type, this research suggests a deep learning-based hybrid model for automated leukemia detection and classification. By taking morphological characteristics and genetic abnormalities into account, leukemic cells have been distinguished. The features in this study are extracted using MobileNetV2, ResNet-101, and VGG-16. Then, feature-level stacking is performed using the Support Vector Machine and Random Forest classifiers for final classification using Principal Component Analysis (PCA). Utilising image segmentation, normalisation, data augmentation, and data oversampling techniques, the pipeline improves data quality and corrects class imbalance. Additionally, t-distributed Stochastic Neighbour Embedding (t-SNE), which shows the extracted features used for the detection of leukemia subtypes, and Gradient-Weighted Class Activation Mapping (Grad-CAM) images help with interpretability by highlighting important decision areas. The suggested study achieved 98.35% accuracy, 95.87% precision, 95.84% sensitivity, 98.97% specificity, and 95.74% F1-Score. Along with the trial results, a comparison of the four separate frameworks, viz., MobileNetV2, ResNet-101, VGG-16, and Vision Transformer, has also been carried out. The comparison shows that the proposed model outperforms the other frameworks. The outcomes show that the suggested model has the capability to be used a reliable means for the prompt identification of AML and its subtypes.

Evaluating and Improving Algorithms for Off-Road Routing

Jay Kansara — 2025-10-25

In this paper, we explore the multifaceted capabilities of route-finding algorithms and their role in delivering dynamic paths for diverse navigation scenarios. In off-road route finding, the shortest path is not always the best; the route's smoothness must also be considered. Present methodologies like A* have notable limitations, such as difficulty adapting to complex terrains and producing rugged routes. This limitation hampers performance in critical scenarios, such as in emergencies like landslides or earthquakes, where off-road exploration is crucial. This paper proposes 3 methodologies that have been fine-tuned with optimal hyperparameters for optimal results using gradient descent. Two of these methodologies, Simulated Annealing and Ant Colony Optimization, overcome some limitations of A* but not all. However, Q-Learning significantly overcomes the limitations and saves travel time by providing a route with 70% fewer undulations and a 16% reduction in route length compared to A*. Compared to other implementations, the Q-Learning implementation proposed in this research not only focuses on minimizing path length but also minimizes route undulations, providing a dual objective approach that is well suited to real-world scenarios. Unlike prior implementations, which focus on a single objective, such as path length or obstacle avoidance, this work leverages a reward function that penalizes elevation variance while rewarding shorter routes, resulting in smoother, more easily traversable paths. Thus, Q-Learning overcomes the cons of the present methodologies and can form synergistic combinations, enhancing the overall performance of off-road space searching systems and accommodating the various challenges and complexities inherent in the targeted applications. The dataset used includes features such as latitude, longitude, and elevation. The strategic application of heuristics enables the swift evaluation of multiple paths, facilitating the selection of optimal routes in real-time applications. By combining various heuristics, the development of off-road path identification systems capable of discerning optimal paths across varied terrains becomes feasible.

A Multi-Label Toxic Comment Classification Framework using under Sampling and Deep Learning Models

Sushma S — 2025-10-27

Toxic online content poses significant challenges to digital communication platforms, necessitating accurate and balanced classification strategies. Unlike traditional binary classification approaches, this study focuses on multi-label toxic comment classification using the Jigsaw dataset, where each comment may exhibit multiple overlapping toxicity types. To address the severe class imbalance inherent in the dataset, three tailored undersampling strategies—One-vs-Rest Undersampling, Multilabel Random Undersampling (MLRU), and an Improved Threshold-Based Undersampling—are proposed to enhance the representation of minority labels such as threat and identity hate. These undersampled datasets are evaluated using a diverse set of ML and DL models, including Random Forest, XGBoost, CNN, RNN, LSTM, BiLSTM, BERT, and RoBERTa. Experimental results shown that this joint multi-label–specific under sampling combined with advanced classification architectures establish superior models, more evident in early detection of rare but relevant types of toxicity. This work demonstrates that multi-label learning frameworks can serve as an effective approach toward fair and full toxic comment detection.

A Systematic-Architectural-Perspective Based Performance Analysis of A-MERIT-C- Dynamic Learning Multitiered Ensemble-Based Real Time Flight Data Analysis

Shailaja B. Jadhav — 2025-10-30

Large-scale data analysis has been the subject of numerous studies recently. In many applications of today's data-intensive world, data is typically brought in continually as data streams. Analytics engines that handle streaming data must be able to react to data that is in motion. Data streams provide special challenges because traditional methods for data mining and machine learning are meant for static information. They are less suited to consider the representative characteristics of data streams and are very less suitable to effectively analyse data that is growing quickly. The authors through this research viz. A-MERIT-C - a dynamic learning multitiered ensemble-based flight real time data analysis system. Through this research authors have presented an active learning dynamic real time data stream analysis model built with self-tuning ensemble learning framework, able to quickly adapt to concepts in near real time streaming data analysis. The conceptual architectural framework illustrated through this research is adaptive to deal with the dynamics related with real time data through the evolving classifier pool (i.e. best performing classifiers get added to classifier pool at every epoch). One more distinguishing characteristic of -A-MERIT-C is instead of using traditional hold out evaluation, it uses prequentially evaluated classifiers. A-MERIT-C's unique features provide significant gains in accuracy, precision, and AUC for streaming data analytics; however, it can also overcome the drawbacks of current algorithms, including concept evolution and feature drift, by using incremental learning and feedback.

Optimized Vision Transformer Architecture for Cardiac Auscultation Classification using GAN augmented MFCC representations

Divya Lalita Sri Jalligampala — 2025-11-04

Heart auscultation is a key diagnostic tool for detecting cardiac abnormalities; however, human interpretation is subjective and prone to error. Classic machine learning algorithms like LSTMs and BiLSTMs have been employed for computer-aided heart sound classification but face challenges with handling acoustic variation, data sparsity, and long-range correlations in spectrograms. Solo Vision Transformers (ViT’s) improve feature extraction but require large datasets to function best. This article introduces a hybrid model combining a Generative Adversarial Network (GAN) and a Vision Transformer (ViT) to address these issues, applying GAN-based data augmentation to enhance training diversity and leveraging ViT's self-attention mechanism to interpret spectrograms better. The data, accessed through the iStethoscope Pro app and clinical testing with DigiScope, comprised normal, murmur, and artifact classes. Preprocessing included silent cutting, resampling, and extraction of MFCCs, spectral contrast, chroma features, and RMSE. The proposed GAN+ViT model was compared to BiLSTM, LSTM, and standalone ViT. The performance showed that GAN+ViT outperformed all baseline models with 90% accuracy, 0.90 F1-score, 0.91 precision, and 0.89 recall, and AUC-ROC values of 0.92 for artifacts, 0.93 for murmurs, and 0.91 for normal sounds. On the other hand, BiLSTM (85%), LSTM (83%), and ViT (80%) were poor in their performance, particularly in discriminating between murmurs and normal sounds. The improved classification power of the hybrid model is due to complementary data augmentation and attention-based feature learning, thereby reducing misclassifications. This research recommends that GAN+ViT is a viable method for automated analysis of cardiac sounds, with high accuracy and generalizability for clinical applications. Future research could explore multimodal integration with ECG data and employ explainable AI methods to enhance diagnostic consistency.

Performance Analysis of Various FACTS Devices Incorporated in Power System Damping Controllers

Chandan Kumar Sah — 2025-11-04

Low-frequency oscillations (LFOs) present a considerable challenge to the long-term viability of contemporary interconnected electrical networks. These oscillations may cause instability in the system, increased equipment stress, and reduced reliability in power delivery. Traditional controllers, such as Power System Stabilizers (PSS), have limitations in efficiently reducing oscillations between areas, particularly in large, multi-machine networks. To address this, Flexible AC Transmission System (FACTS) devices present a promising alternative due to their ability to dynamically control transmission line parameters and enhance system stability. This research proposes an integrated damping strategy using four FACTS controllers ‘Static VAR Compensator (SVC), Unified Power Flow Controller (UPFC), Static Synchronous Compensator (STATCOM), and Static Synchronous Series Compensator (SSSC) for effective suppression of LFOs. The controllers settings are adjusted for maximum effectiveness utilizing the Ant Colony Optimization (ACO) algorithm to adapt to changing system conditions in real-time. The proposed control approach is tested within the IEEE 30-bus system under normal, light, and heavy loading conditions. The simulation outcomes demonstrate that ACO-optimized FACTS controllers significantly outperform both the base case and non-optimized configurations by reducing settling time, overshoot, and power losses, thereby enhancing overall system stability and damping performance. Performance is measured using indices like Integral of Squared Error (ISE), Integral of Time Squared Error (ITSE), Integral of Absolute Error (IAE), Integral of Time Absolute Error (ITAE), LMN (Line Stability Index), FVSI (Fast Voltage Stability Index) and VSF (Voltage Sensitivity Factor).

Triple Attention-Based Hybrid Deep Learning Framework for Enhanced Stock Market Prediction

Pranjali Kasture — 2025-11-06

Stock price prediction is a complex problem because financial time series data are volatile and complicated. The model should learn the temporal relationship and complex spatial patterns in data for precise stock price prediction. Conventional methods used for stock price forecasting have many limitations regarding handling nonlinear, complex, and dynamic data. This study assesses a hybrid deep learning model integrated with a triple attention mechanism to predict stock prices. It is experimental that the proposed MTA-HDCRNN model performs well on intricate data. The deep CNN works well on finding the local patterns in the data, whereas the simple RNN supports to learn sequential data. The triple attention mechanism emphasizes which features to focus on and where to focus. The dataset used for analysis is the BSE and Nifty 50. Web scraping is done to get the news data. Feature extraction includes statistical features, entropy features, PCA features, and technical indicators. Overall, the complete architecture of the proposed model is vigorous. It is observed that there is a 2% to 6% decrease in error values when the model is compared with existing state-of-the-art models. Experimentation shows that the proposed model enhances the stock price prediction, making it useful for investors and financial analysts for decision-making.

Comprehensive Analysis of Structural and Magnetic Properties of Mn2CoFeGe2 Double Half-Heusler Alloy

Naaraayanan T — 2025-11-10

Heusler alloys have unusual properties, making them highly interesting for evaluation. This study focuses on the double half-Heusler Mn₂CoFeGe_2, analyzing its crystal structure and magnetic properties. The prepared alloy was analysed through X-ray diffraction to study its structural features. The material’s morphology was examined using scanning electron microscopy. Simultaneously, the elemental constituent proportions of Mn₂CoFeGe₂ were confirmed using energy-dispersive X-ray spectroscopy (EDS) analysis, which verifies that the elements are present in a 2:1:1:2 atomic ratio. Interestingly, the saturation magnetization slightly decreases from 60 emu/g at 5 K to 57 emu/g at 150 K. The measured g-value of the alloy is greater than 2. The Mn₂CoFeGe₂ alloy behaves like a semiconductor based on UV measurements. Therefore, the use of this alloy in the new technologies seems reasonable.