Predictive Framework for Sustainable Engineering through Machine Learning and Cross-Sector Collaboration Abhik Choudhary, Upasana Adhikari, Dipankar Roy, Subir Gupta, Priyanka Roy, Aparna Bhaduri EPJ Web of Conferences, 2025 Accomplishing Sustainable engineering rests on careful collaboration across disciplines to solve challenges of the modern world; this includes climate change, resource management, and socio-economic issues. Still, collaboration across the different sectors of AIG (academia, industry, and government) is poorly integrated due to operational silos and structures lacking a centralized data-driven approach. This research proposes a new methodology based on ensemble machine learning, assessing, and predicting the outcomes of engineering projects with real-time open-source data focused on sustainability. Unlike traditional rule-based and qualitative evaluations, the proposed method measures diverse parameters like funding streams, policy advocacy, and stakeholder participation to construct a comprehensive model of collaboration at cross-sectoral levels. Harvested public data was normalized, encoded, and pre-processed class balances using SMOTE. An Extra Trees Classifier was trained to perform binary classification of project success deemed as primary indicators, evaluating performance through accuracy, precision, recall, F1 score, log loss, and MSE. The model achieved 96% accuracy and 0.733 F1 score. Feature importance analysis corroborated interpretability of model predictions. These results underscored the drivers of effective collaboration while showcasing the model's robust predictive capacity. Beyond bridging the literature gap, this study equips policymakers and other stakeholders with actionable insights to enhance strategic planning, resource distribution, and governance in sustainable engineering.
Predicting Social Media Virality using Stacking Ensemble with Random Forest, XGBoost and Logistic Regression Upasana Adhikari, Subir Gupta, Joyjit Patra Proceedings of 8th International Conference on Inventive Computation Technologies Icict 2025, 2025 The research investigates predicting the virality of content on social media platforms using a stacking ensemble model with Random Forest, XGBoost, and Logistic Regression. Social media trends and user interactions remain largely uncontrollable, but this approach seeks to address those issues using sophisticated machine learning methods. Parlor models in itchnology tend to ignore the persistent inconsistencies associated with worldwide phenomena of viral content or fail to generalize adequately. The study model addresses these shortcomings with a balanced dataset and attention to feature importance calculated with user engagement defined as shares, likes, and comments. Incorporating Random Forest and XGBoost improves model robustness when faced with complicated cases while increasing overall accuracy. At the same time, “meta-classifier” Logistic Regression further improves results by integrating diverse base model predictions. This research is unique in employing an optimized stacking ensemble model to significantly enhance predictions of social media virality. This model extends academic frontiers and provides actionable knowledge to firms and content developers who wish to improve their engagement through targeted marketing in the ever-evolving digital ecosystem.
Computational Analysis of Genre Effects on Movie Ratings Using MLP Algorithms Subir Gupta, Upasana Adhikari, Shefali Varshney, Tanupriya Choudhury Journal of Computer Science, 2025 Predictive analytics are what the entertainment industry depends on so much in prognosticating movie ratings, which is why they inform filmmakers, distributors and stream platforms strategic moves. Traditional prediction models most times only succeed in missing out on the intricate dynamics that are genre-specific to movie ratings thereby leading to inaccuracies and suboptimal decision making. The paper proposes this research presents a comprehensive mechanism that combines extensive metadata with Multi-Layer Perceptron (MLP) models to increase the accuracy of predictions across multiple cinematic genres. Therefore, we had a goal of establishing fine patterns due to MLP regressors based on specific genres as well as addressing limitations linked to traditional approaches. To conduct this study; we used principal component analysis and one-hot encoding for 950 films followed by genre-specific modeling alongside statistical tests such as ANOVA, t-tests, Gradient Boosting Classifiers among others for model validation. They found that adventure movies were more predictable than other genres (MSE = 0.023) such as action (MSE = 2.816). It’s clear then that accurate modelling requires an examination by gender and broad data sources integration. The research emphasizes the potential of improved machine learning methods to change predictive modeling in the area of art. Further work will seek to develop more accurate feature selection, deal with data imbalance and incorporate real-time audience engagement measures into the optimization process for better predictions that would help film makers make better strategic decisions.
Harnessing Adaptive Intelligence for Strategic Human Capital Retention Niloy Kumar Bhattacherjee, Sayanti Samanta, Indrani Sengupta, Sandip Mukherjee, Upasana Adhikari, Subir Gupta International Conference on Computing Intelligence and Application Ciacon 2025, 2025 The productivity of an organization is negatively impacted by employee attrition, which increases operational costs and hinders the effectiveness of human resource (HR) management systems. To explore this seemingly unresolvable problem, the study introduces a new solution using the TabNet deep learning model because it suits tabular data and provides both accuracy and interpretability. Unlike traditional models which require extensive manual feature engineering where low transparency renders the model impractical for HR use, TabNet does not suffer from such limitations. The model’s sequential attention mechanism enables dynamic feature selection along with offering insight into the prediction process while still retaining predictive performance. This research seeks to predict employee retention using HR records that include demographic information, job satisfaction levels, and work attendance patterns. Standard performance evaluation metrics were used to evaluate the TabNet model, including accuracy, F1 score, precision, recall, and log loss. The results were exceptional, achieving 97.2% accuracy, 98.1% F1 score, 99.0% recall, and 1.47 log loss outperforming traditional machine learning techniques. The results demonstrate the model’s ability to help HR departments shift from reactive to proactive workforce management, accurately predicting high-risk cases of attrition and aiding in the formulation of retention strategies, validating the model's ability.
Prediction of Social Media Virality Through Stacking Ensemble utilising ML Algorithms Upasana Adhikari, Subir Gupta, Joyjit Patra, Bibhuti Bhusan Dash, Subrata Chowdhury, Sudhansu Shekhar Patra 2025 6th International Conference for Emerging Technology Incet 2025, 2025 The user interaction datasets are too complex with a lot of noise which makes predicting virality with social media content extremely difficult. This multi-faceted problem requires different models to come up with a solution, however, most single-model approaches are proving to be inefficient due to the lack of accuracy and reliability in predictive performance. This research comes up with a new stacking ensemble model that incorporates Random Forest, XGBoost, and Logistic Regression, which aims to solve the problem stated before. The main focus is improving model robustness while retaining a clear interpretive framework and incorporating weakly labeled dataset interpretability. Employing bagging, boosting, and probabilistic calibration simultaneously improves accuracy and reliability without sacrificing ensemble bias. The model was viral content detected at 95% accuracy, 98% F1-score, with 90% precision and recall, and 92% ROC-AUC showcasing excellent metrics predictive performance. Improved prediction reliability was ensured through calibrated outputs alignment by regression with real-world distribution alignment enhancing predictive dependability. This model provides strong efficient scalable structure to prediction virality and emphasizes its potential across diverse data-centric domains.
Optimising Energy Efficiency in Agricultural Consumer Electronics Using Principal Component Analysis and Deep Q-Learning Subir Gupta, Upasana Adhikari, Pinky Pramanik, Subrata Chowdhury, Shreyas J., Anurag Sinha, Saifullah Khalid, Malathi S. Y. Iet Cyber Physical Systems Theory and Applications, 2025 The ability to reduce emissions and improve sustainability in agricultural consumer electronics has been significantly hindered due to the use of energy‐intensive technology within the agricultural sector. This study proposes a new enhancement of deep Q‐learning (DQN) with principal component analysis (PCA) focused on energy efficiency. PCA helps manage massive operational data by performing dimensionality reduction, whereas DQN, a reinforcement learning paradigm, optimises decision‐making during real‐world interactions. The main contribution of this study is in the combined use of PCA and DQN to form customisable, precise, contest‐responsive energy frameworks powered by real‐time analytics on agricultural data—energy management on such a scale has not been approached in the context of sustainable agriculture before. The experiments confirm the optimal model, further achieving a cumulative reward of 72.56, an average emission of 1.83, a Q ‐value of 24.76 and a total zenith value of 75.40% in ensuring numerous noncriteria‐defined efficient energy‐dependent operations. This paradigm not only fills the void in the automation of passive intelligent agricultural systems but also serves as a point of reference for other eco‐critical domains to strive towards greener technology.
Predictive Modeling of Tiger Population Trends in Indian Reserves Mansi Prabha, Arya Raushan, M. D. Nasrullah, Upasana Adhikari, Subir Gupta, Subrata Chowdhury, Thu Thi Nguyen, Duc-Tan Tran 2024 International Conference on Decision Aid Sciences and Applications Dasa 2024, 2024 The goal of this study is to develop predictive models to forecast future tiger population trends in protected reserves across India, with a focus on providing actionable conservation strategies. Utilizing a combination of time series analysis and regression techniques, specifically the ARIMA model, the research aims to assess the impact of environmental factors such as habitat quality, prey availability, and human disturbances on tiger populations. The methodology involves collecting historical tiger population data from wildlife agencies, followed by data cleaning and exploratory analysis to identify key trends. The ARIMA model will predict population trends, while regression models will evaluate the effects of environmental variables. The models' accuracy will be assessed using performance metrics such as RMSE and MAE, with cross-validation to ensure generalization. This comprehensive approach aims to provide wildlife managers with data-driven recommendations for targeted interventions, such as improving habitat connectivity or controlling human encroachment. The study's novel integration of multiple environmental factors into a holistic predictive model distinguishes it from previous research, offering a robust tool for tiger conservation efforts.
RECENT SCHOLAR PUBLICATIONS
Dynamic Classification of Purchasing Patterns using Simplified Q-learning S Gupta, M Chakraborty, U Adhikari, BB Dash, UC De, SS Patra 2025 3rd International Conference on Sustainable Computing and Smart Systems … , 2025 2025
Harnessing Adaptive Intelligence for Strategic Human Capital Retention NK Bhattacherjee, S Samanta, I Sengupta, S Mukherjee, U Adhikari, ... 2025 International Conference on Computing, Intelligence, and Application … , 2025 2025
Enhancing Text Intelligence with Soft Voting and TF-IDF Logistic Learners P Pramanik, S Samanta, RK Mondal, J Patra, U Adhikari, S Gupta International Conference on Data Analytics & Management, 301-313 , 2025 2025 Citations: 1
Prediction of Social Media Virality Through Stacking Ensemble utilising ML Algorithms U Adhikari, S Gupta, J Patra, BB Dash, S Chowdhury, SS Patra 2025 6th International Conference for Emerging Technology (INCET), 1-5 , 2025 2025 Citations: 2
IoT-Integrated reinforcement Learning-Based mine detection system for military and humanitarian applications S Gupta, U Adhikari, D Roy, S Hazra ICCK Transactions on Machine Intelligence 1 (1), 17-28 , 2025 2025 Citations: 8
Predicting Social Media Virality Using Stacking Ensemble with Random Forest, XGBoost and Logistic Regression U Adhikari, S Gupta, J Patra 2025 International Conference on Inventive Computation Technologies (ICICT … , 2025 2025
Computational Analysis of Genre Effects on Movie Ratings Using MLP Algorithms S Gupta, U Adhikari, S Varshney, T Choudhury Journal of Computer Science 21 (4) , 2025 2025 Citations: 2
Optimising Energy Efficiency in Agricultural Consumer Electronics Using Principal Component Analysis and Deep Q‐Learning S Gupta, U Adhikari, P Pramanik, S Chowdhury, A Sinha, S Khalid, M SY IET Cyber‐Physical Systems: Theory & Applications 10 (1), e70029 , 2025 2025 Citations: 1
Predictive Framework for Sustainable Engineering through Machine Learning and Cross-Sector Collaboration A Choudhary, U Adhikari, D Roy, S Gupta, P Roy, A Bhaduri EPJ Web of Conferences 328, 01048 , 2025 2025
Predictive Modeling of Tiger Population Trends in Indian Reserves M Prabha, A Raushan, MD Nasrullah, U Adhikari, S Gupta, S Chowdhury, ... 2024 International Conference on Decision Aid Sciences and Applications … , 2024 2024 Citations: 2
MOST CITED SCHOLAR PUBLICATIONS
IoT-Integrated reinforcement Learning-Based mine detection system for military and humanitarian applications S Gupta, U Adhikari, D Roy, S Hazra ICCK Transactions on Machine Intelligence 1 (1), 17-28 , 2025 2025 Citations: 8
Prediction of Social Media Virality Through Stacking Ensemble utilising ML Algorithms U Adhikari, S Gupta, J Patra, BB Dash, S Chowdhury, SS Patra 2025 6th International Conference for Emerging Technology (INCET), 1-5 , 2025 2025 Citations: 2
Computational Analysis of Genre Effects on Movie Ratings Using MLP Algorithms S Gupta, U Adhikari, S Varshney, T Choudhury Journal of Computer Science 21 (4) , 2025 2025 Citations: 2
Predictive Modeling of Tiger Population Trends in Indian Reserves M Prabha, A Raushan, MD Nasrullah, U Adhikari, S Gupta, S Chowdhury, ... 2024 International Conference on Decision Aid Sciences and Applications … , 2024 2024 Citations: 2
Enhancing Text Intelligence with Soft Voting and TF-IDF Logistic Learners P Pramanik, S Samanta, RK Mondal, J Patra, U Adhikari, S Gupta International Conference on Data Analytics & Management, 301-313 , 2025 2025 Citations: 1
Optimising Energy Efficiency in Agricultural Consumer Electronics Using Principal Component Analysis and Deep Q‐Learning S Gupta, U Adhikari, P Pramanik, S Chowdhury, A Sinha, S Khalid, M SY IET Cyber‐Physical Systems: Theory & Applications 10 (1), e70029 , 2025 2025 Citations: 1
Dynamic Classification of Purchasing Patterns using Simplified Q-learning S Gupta, M Chakraborty, U Adhikari, BB Dash, UC De, SS Patra 2025 3rd International Conference on Sustainable Computing and Smart Systems … , 2025 2025
Harnessing Adaptive Intelligence for Strategic Human Capital Retention NK Bhattacherjee, S Samanta, I Sengupta, S Mukherjee, U Adhikari, ... 2025 International Conference on Computing, Intelligence, and Application … , 2025 2025
Predicting Social Media Virality Using Stacking Ensemble with Random Forest, XGBoost and Logistic Regression U Adhikari, S Gupta, J Patra 2025 International Conference on Inventive Computation Technologies (ICICT … , 2025 2025
Predictive Framework for Sustainable Engineering through Machine Learning and Cross-Sector Collaboration A Choudhary, U Adhikari, D Roy, S Gupta, P Roy, A Bhaduri EPJ Web of Conferences 328, 01048 , 2025 2025
