Template-Type: ReDIF-Article 1.0 Author-Name: Jayanta Mondal Author-X-Name-First: Jayanta Author-X-Name-Last: Mondal Author-Name: Jitendra Pramanik Author-X-Name-First: Jitendra Author-X-Name-Last: Pramanik Author-Name: Satyajit Pattnaik Author-X-Name-First: Satyajit Author-X-Name-Last: Pattnaik Author-Name: Bijay Kumar Paikaray Author-X-Name-First: Bijay Kumar Author-X-Name-Last: Paikaray Title: Using the BIRCH algorithm and affinity propagation, an advanced descriptor for video processing Abstract: Video summarisation provides concise and non-redundant information for object and intrusion detection in video surveillance. As video content continues to expand quickly, an automatic video summary would be helpful for anyone who wants to learn more quickly and with less effort. Most existing methods depend on various network architectures to train a single score predictor for shot rating and selection. This study addresses the issue of video summarisation, which involves selecting significant frames to succinctly and comprehensively express the original film's material. The current paper presents a comparative study of the application of advanced texture descriptors local phase quantisation (LPQ), local ternary pattern (LTP), and local binary pattern (LBP) in the process of video summarisation. Clusters of keyframes have been extracted by unsupervised learning algorithms - Affinity Propagation % BIRCH. The performance of the proposed video summarising method has shown good trial results. Journal: Int. J. of Data Analysis Techniques and Strategies Pages: 25-40 Issue: 1 Volume: 18 Year: 2026 Keywords: LTP; local ternary pattern; LBP; local binary pattern; affinity propagation; LPQ; local phase quantisation; BIRCH; key feature. File-URL: http://www.inderscience.com/link.php?id=151636 File-Format: text/html File-Restriction: Access to full text is restricted to subscribers. Handle: RePEc:ids:injdan:v:18:y:2026:i:1:p:25-40 Template-Type: ReDIF-Article 1.0 Author-Name: Rohit Rastogi Author-X-Name-First: Rohit Author-X-Name-Last: Rastogi Author-Name: Saransh Chauhan Author-X-Name-First: Saransh Author-X-Name-Last: Chauhan Author-Name: Yash Rastogi Author-X-Name-First: Yash Author-X-Name-Last: Rastogi Author-Name: Vaibhav Aggarwal Author-X-Name-First: Vaibhav Author-X-Name-Last: Aggarwal Author-Name: Utkarsh Agrawal Author-X-Name-First: Utkarsh Author-X-Name-Last: Agrawal Author-Name: Richa Singh Author-X-Name-First: Richa Author-X-Name-Last: Singh Title: Prediction model for AQI through Indian Vedic science: knowledge management technique to control pollution and for sustainable society Abstract: The paper provides an essence of how Indian Vedic Sciences can be used for preventing and predicting the ill effects of pollution on the human body and nature through adopting simple methods of Yajna and Hawan in daily routine. With respect to any other resource like land and water, air is considered as the most important resource. Evidence shows that Indian Vedic Sciences primarily focus on 'prana vayu' which means 'air that we breathe'. The author's team and the Central Pollution Control Board (CPCB) have gathered the data and reading of the last four months through installed sensors in an isolated as well as non-isolated environment that was continuously under the effects of Yajna and Hawan. Journal: Int. J. of Data Analysis Techniques and Strategies Pages: 82-106 Issue: 1 Volume: 18 Year: 2026 Keywords: AQI; air quality index; PM 2.5; PM 10; climate change; Yajna; mantra; knowledge management; sustainable society; knowledge levels and extractions. File-URL: http://www.inderscience.com/link.php?id=151637 File-Format: text/html File-Restriction: Access to full text is restricted to subscribers. Handle: RePEc:ids:injdan:v:18:y:2026:i:1:p:82-106 Template-Type: ReDIF-Article 1.0 Author-Name: Emir Mahmood Kalik Author-X-Name-First: Emir Mahmood Author-X-Name-Last: Kalik Author-Name: Ayad Hasan Adhab Author-X-Name-First: Ayad Hasan Author-X-Name-Last: Adhab Title: Design a modern scheme for machine learning-based detection of image forgery Abstract: The rapid growth and development of information technology have led to the emergence of numerous methods that are used for digital image forgery. Thus, manipulating digital images to achieve a negative or positive purpose has become easy. The use of advanced methods in forgery has increased the difficulty of detecting the nature of the images, whether they are original or forged, especially when using classical methods. Therefore, many researchers are interested in this field, making it a popular research direction for researchers. In this paper, we will introduce an intelligent approach to designing a method for digital image forgery detection by using machine learning. This proposal seeks to train an intelligent model to discern between altered and original images by examining the essential features of the images. The results demonstrated that it achieved superior performance and high accuracy when it came to detecting forgeries in digital images. Journal: Int. J. of Data Analysis Techniques and Strategies Pages: 41-56 Issue: 1 Volume: 18 Year: 2026 Keywords: CNN; convolutional neural network; DRL; deep reinforcement learning; forgery; image detection; manipulation. File-URL: http://www.inderscience.com/link.php?id=151638 File-Format: text/html File-Restriction: Access to full text is restricted to subscribers. Handle: RePEc:ids:injdan:v:18:y:2026:i:1:p:41-56 Template-Type: ReDIF-Article 1.0 Author-Name: Anton Abdulbasah Kamil Author-X-Name-First: Anton Abdulbasah Author-X-Name-Last: Kamil Author-Name: Muhamad Safiih Lola Author-X-Name-First: Muhamad Safiih Author-X-Name-Last: Lola Title: Development of G-causality by utilising hybridisation of bootstrap method for assessing tourism impacts in Malaysia Abstract: This study aims to develop and examine the causality direction of non-economic short and long-term factors in the Malaysian tourism industry using a new hybrid Bootstrap-Granger Model. The proposed method was validated with non-economic factor dataset from the World Bank (tourist arrival, population, air transport, and carbon dioxide emission) in the tourism industry. The model effectiveness was tested and analysed by comparing it against the actual Granger model using statistical tests such as unit root, Johansen cointegration, and Granger causality tests. The empirical results revealed that compared to the Granger model, the proposed counterpart generated smaller mean square error and root mean square error values for non-economic factor datasets. Furthermore, the results also revealed that tourist arrival and other determinants were co-integrated. In other words, the proposed model enhanced Granger causality accuracy and proved to be more robust, precise, and accurate results towards the promotion of overall economic activities. Journal: Int. J. of Data Analysis Techniques and Strategies Pages: 57-81 Issue: 1 Volume: 18 Year: 2026 Keywords: bootstrap method; Granger Causality; hybridisation; tourism impact and non-economy factors; Malaysia. File-URL: http://www.inderscience.com/link.php?id=151639 File-Format: text/html File-Restriction: Access to full text is restricted to subscribers. Handle: RePEc:ids:injdan:v:18:y:2026:i:1:p:57-81 Template-Type: ReDIF-Article 1.0 Author-Name: Hoang Duc Le Author-X-Name-First: Hoang Duc Author-X-Name-Last: Le Author-Name: Ke Nghia Nguyen Author-X-Name-First: Ke Nghia Author-X-Name-Last: Nguyen Title: Application of generalised regression neural network for financial time series forecasting: a comprehensive comparison with autoregressive integrated moving average Abstract: Time series forecasting plays a crucial role in fields such as economics, business, and finance. Traditional models like the autoregressive integrated moving average (ARIMA) have been widely used for their accuracy. However, advances in computing and the rise of machine learning (ML) and deep learning (DL) have introduced powerful alternatives. This study examines the performance of a DL-based method - the generalised regression neural network (GRNN) - compared to ARIMA. Results show that GRNN significantly outperforms ARIMA in forecasting accuracy, with an error margin of less than 5%. GRNN also achieves better results across statistical metrics, including mean absolute error (MAE), root mean square error (RMSE), and mean absolute percentage error (MAPE). Additionally, GRNN offers faster training times, making it especially advantageous in scenarios requiring rapid and frequent forecasts. These findings highlight GRNN's potential as a superior tool for time series prediction in dynamic, data-driven environments. Journal: Int. J. of Data Analysis Techniques and Strategies Pages: 1-24 Issue: 1 Volume: 18 Year: 2026 Keywords: time series forecasting; machine learning; deep learning; GRNN; generalised regression neural network; ARIMA; autoregressive integrated moving average. File-URL: http://www.inderscience.com/link.php?id=151640 File-Format: text/html File-Restriction: Access to full text is restricted to subscribers. Handle: RePEc:ids:injdan:v:18:y:2026:i:1:p:1-24