UAUpasana Adhikari

Upasana Adhikari

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Upasana Adhikari has a total of 12 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Enhancing Text Intelligence with Soft Voting and TF-IDF Logistic Learners

    by Pinky Pramanik, Sayanti Samanta, Rajib Kumar Mondal, Joyjit Patra, Upasana Adhikari, Subir Gupta

  • Open Access

    IoT-Integrated Reinforcement Learning-Based Mine Detection System for Military and Humanitarian Applications

    by Subir Gupta, Upasana Adhikari, Dipankar Roy, Sudipta Hazra

    This research proposes an advanced system for landmine detection combining the internet of things and reinforcement learning, which seeks to resolve issues in conventional methods that misidentify more than 30% of detections, have slow reaction times, and are not suited for different environments. Others like metallic detectors and sniffer dogs also pose greater danger for wrong threat identification, more so due to sluggish attempts. The system proposed in this study is novel in that it customizes metal detection by integrating a sensor into military boots, thus permitting constant scanning without the use of hands. A meta-learning-based Machine Learning model improves detection accuracy. It was found that reward-driven reinforcement learning regulation improves mine detection accuracy, increases the analysis attempts in each evaluation phase, and alters the strategic settings. The range of analysis conducted during this study validates the argument in question but this reworking of the system does not fully resolve the issues. The innovation is having that with proper situational awareness this model enables real-time implementation of IoT devices. This adaptable system is not only advantageous for military endeavors but can also be useful for demining activities. More robust multisensory capabilities are essential to facilitate effective and safe landmine inspection all over the globe, so follow-up studies should concentrate on field trials with accompanying iterative improvements.


  • Predicting Social Media Virality Using Stacking Ensemble with Random Forest, XGBoost and Logistic Regression

    by Upasana Adhikari, Subir Gupta, Joyjit Patra

    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.


  • Open Access

    Optimising Energy Efficiency in Agricultural Consumer Electronics Using Principal Component Analysis and Deep Q‐Learning

    by Subir Gupta, Upasana Adhikari, Pinky Pramanik, Subrata Chowdhury, Shreyas J., Anurag Sinha, Saifullah Khalid, Malathi S. Y.

    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.