Skip to research
Samir Poudel · Portfolio
Research notebookMiddle Tennessee State University

Distributed AI · Privacy · Security

Research & talks.

PhD student in Computational and Data Science

Modern AI systems are getting more powerful and more centralized at the same time. That creates real problems: privacy risks, single points of failure, and data that organizations cannot share. I build distributed AI frameworks that let multiple parties train models together without exposing raw data. My work covers healthcare, cybersecurity, and environmental monitoring.

Browse publications ↓Google Scholar ↗Engineering projects ↗

  • Federated Learning & Privacy-Preserving AI

    I build federated learning frameworks that train models across distributed data owners without centralizing any sensitive data. The main failure modes I address are backdoor attacks, heterogeneous data distributions, and communication bottlenecks that block real-world deployment.

    Related papers
    • RBBD: A Representation-Based Framework for Edge-Case Backdoor Defense in Federated Learning
    • Privacy Meets Conservation: Federated Learning's Revolution in Deforestation Detection

  • Blockchain Security & Smart Contract Analysis

    Smart contracts hold billions of dollars in value but break under subtle code vulnerabilities. I build multimodal tools that fuse source code, opcode, and bytecode to catch vulnerabilities that single-modality approaches miss.

    Related papers
    • VulnFusion: Exploiting Multimodal Representations for Advanced Smart Contract Vulnerability Detection
    • QuadraCode AI: Smart Contract Vulnerability Detection with Multimodal Representation
    • A Blockchain and IoT-Enabled Framework for Ethical and Secure Coffee Supply Chains

  • Robust Machine Learning for Healthcare & Science

    I apply machine learning to safety-critical domains where reliability directly affects patient outcomes. My work covers brain tumor segmentation, Alzheimer's classification, ICU outcome prediction, cervical cancer detection, and physiological stress analysis.

    Related papers
    • Prediction of Length-of-Stay at ICU Using Machine Learning based on MIMIC-III Database
    • Mental Workload Classification from fNIRS Signals by Leveraging Machine Learning
    • Improved Classification of Alzheimer's Disease with Convolutional Neural Networks
    • Brain Tumor Segmentation and Classification Using ACGAN with U-Net and Independent CNN-Based Validation
    • Scalable Multimodal Machine Learning for Cervical Cancer Detection
    • Personalized Stress Detection from Chest-Worn Sensors by Leveraging Machine Learning

  • Distributed Systems & Big Data Engineering

    Scalable AI needs scalable infrastructure. I benchmark distributed data pipelines, optimize parallelization for preprocessing bottlenecks, and build cloud-native ML systems with measurable throughput and latency gains.

    Related papers
    • Minimization of High Computational Cost in Data Preprocessing with MPI4Py
    • A Big Data Optimization Comparison using Apache Spark and Apache Airflow

  • Game Theory & Strategic AI

    Federated learning and multi-agent systems are strategic environments. Participants can act selfishly, collude, or drop out to maximize their own gain. I apply mechanism design, Nash equilibria, and cooperative game theory to build distributed AI systems where honest participation is the rational choice.

    Related papers
    Future directions

    • Federated Foundation Models

      Applying federated learning to large language and vision-language models, where communication costs and data heterogeneity are far greater than in standard FL.

    • Stronger Privacy Guarantees

      Building tighter, composable privacy accounting methods that give real guarantees for deployed federated systems without the accuracy loss of standard DP-SGD.

    • Robust Smart Contract Auditing

      Combining static analysis, fuzzing, and learned representations to build audit pipelines that catch new vulnerability classes as Solidity and the EVM evolve.

    • Privacy-Preserving Environmental AI

      Using federated multi-source remote sensing for climate and biodiversity monitoring, enabling global environmental models without centralizing sovereign satellite data.

    Research, in practiceInteractive figure

    Learning across borders. Keeping imagery local.

    Federated learning · Computer vision · U-Net

    CLIENT 01Data stays localCLIENT 02Data stays localCLIENT 03Data stays localAGGREGATESHARED MODELMODEL UPDATESCollaborative learning

    Each participating site trains on its own satellite imagery.

    97.3%less cross-border data transfer

    Federated deforestation detection achieved 0.85 intersection over union (IoU) on satellite imagery while reducing cross-border data transfer by 97.3%. Explore the code ↗

    Conceptual illustration; three clients shown for clarity. Results are from the experiments reported in my resume, not a live simulation.
    01

    Publications

    Papers and collaborative work.

    1. 2025 · 2025 IEEE 7th International Conference on Trust, Privacy and Security in Intelligent Systems

      RBBD: A Representation-Based Framework for Edge-Case Backdoor Defense in Federated Learning

    2. 2025 · Future Internet 17 (8), 334

      A Blockchain and IoT-Enabled Framework for Ethical and Secure Coffee Supply Chains

    3. 2025 · 2025 IEEE World AI IoT Congress (AIIoT)

      Privacy Meets Conservation: Federated Learning's Revolution in Deforestation Detection

    4. 2025 · 2025 IEEE 15th Annual Computing and Communication Workshop (CCWC)

      A Big Data Optimization Comparison using Apache Spark and Apache Airflow

    5. 2025 · 2025 IEEE 15th Annual Computing and Communication Workshop (CCWC)

      Chain Your Loot: Implementing Blockchain into Gaming Loot Box Markets

    6. 2024 · 2024 IEEE Signal Processing in Medicine and Biology Symposium (SPMB)

      Brain Tumor Segmentation and Classification Using ACGAN with U-Net and Independent CNN-Based Validation

    7. 2024 · 2024 6th International Conference on Blockchain Computing and Applications (BCCA)

      VulnFusion: Exploiting Multimodal Representations for Advanced Smart Contract Vulnerability Detection

    8. 2024 · 2024 33rd International Conference on Computer Communications and Networks (ICCCN)

      QuadraCode AI: Smart Contract Vulnerability Detection with Multimodal Representation

    9. 2024 · 2024 IEEE World AI IoT Congress (AIIoT)

      Scalable Multimodal Machine Learning for Cervical Cancer Detection

    10. 2024 · 2024 International Conference on Electrical, Computer and Energy Technologies (INTCEC)

      Personalized Stress Detection from Chest-Worn Sensors by Leveraging Machine Learning

    11. 2024 · International Conference on Advances in Computing Research

      Credit Card Batch Processing in Banking System

    12. 2023 · 2023 IEEE International Conference on Big Data (BigData)

      Using the CARLA Simulator to Train a Deep Q Self-Driving Car

    13. 2023 · 2023 IEEE Signal Processing in Medicine and Biology Symposium (SPMB)

      Mental Workload Classification from fNIRS Signals by Leveraging Machine Learning

    14. 2023 · 2023 IEEE Signal Processing in Medicine and Biology Symposium (SPMB)

      Improved Classification of Alzheimer's Disease with Convolutional Neural Networks

    15. 2023 · 2023 IEEE Conference on Artificial Intelligence (CAI)

      Prediction of Length-of-Stay at ICU Using Machine Learning based on MIMIC-III Database

    16. 2023 · Journal of Marketing Analytics

      Cryptocurrency Price and Volatility Predictions with Machine Learning

    17. 2023 · Machine Learning with Applications

      Minimization of High Computational Cost in Data Preprocessing with MPI4Py

    02

    Talks

    Conference presentations and research conversations.

    • 2025 · IEEE TPS / CIC 2025 · Pittsburgh, PA

      RBBD: A Representation-Based Framework for Edge-Case Backdoor Defense in Federated Learning

    • 2025 · IEEE AIIoT 2025 · Seattle, WA

      Privacy Meets Conservation: Federated Learning's Revolution in Deforestation Detection

    • 2025 · IEEE CCWC 2025 · Las Vegas, NV

      A Big Data Optimization Comparison Using Apache Spark and Apache Airflow

    • 2025 · IEEE CCWC 2025 · Las Vegas, NV

      Chain Your Loot: Implementing Blockchain Into Gaming Loot Box Markets

    • 2024 · ICCCN 2024

      QuadraCode AI: Smart Contract Vulnerability Detection with Multimodal Representation

    • 2024 · IEEE AIIoT 2024 · Seattle, WA

      Scalable Multimodal Machine Learning for Cervical Cancer Detection

    • 2024 · ACR 2024 · Madrid, Spain

      Comparative Study of Real-Time and Batch Processing Approaches in ML-Based Fraud Detection

    • 2023 · IEEE Big Data 2023 · Sorrento, Italy

      Using the CARLA Simulator to Train A Deep Q Self-Driving Car

    • 2023 · ACM Mid-Southeast 2023 · Gatlinburg, TN

      GAN-Based Data Augmentation for Chest X-ray Classification in Lung Disease Diagnosis

    • 2023 · ACM Mid-Southeast 2023 · Gatlinburg, TN

      Empowering AI's Trust and Big Data Resources in Healthcare Industry with Blockchain

    • 2023 · IEEE SPMB 2023

      Improved Prediction of Classification of the Alzheimer's Disease with Convolutional Neural Network

    03

    Figures

    A closer look at the work.

    Loading figures…