About me

I am a PhD student in the Department of Electronics and Telecommunications at Politecnico di Torino, supervised by Prof. Cristina Rottondi and Prof. Andrea Bianco.

My current research focuses on semantic communication systems, which transmit the meaning of information rather than raw signal, using deep learning models trained to account for real optical fiber channel effects, including scenarios that span multiple optical bands. This sits alongside ongoing work on reliability and fairness in machine learning for optical networks, covering bias mitigation in models that estimate quality of transmission for optical lightpaths, and risk aware classification of soft failures in optical networks.

Together, these directions explore how machine learning for optical networks can become more reliable and fair, and how transmission itself can move beyond raw signal toward meaning.

What i'm doing

  • semantic communications icon

    Semantic Communications

    Deep-learning joint source–channel coding that transmits task-relevant meaning over multi-band optical fiber channels.

  • bias mitigation icon

    Bias Mitigation in ML

    Reducing bias in models that estimate the quality of transmission of optical lightpaths.

  • soft failure classification icon

    Soft Failure Classification

    Risk-aware models for detecting and classifying soft failures in optical networks.

  • open source projects icon

    Open Source Projects

    Building and contributing to open-source projects, including a stock-market analysis tool on GitHub.

Daniel lewis
quote icon

Daniel lewis

Richard was hired to create a corporate identity. We were very pleased with the work done. She has a lot of experience and is very concerned about the needs of client. Lorem ipsum dolor sit amet, ullamcous cididt consectetur adipiscing elit, seds do et eiusmod tempor incididunt ut laborels dolore magnarels alia.

Resume

Education

  1. Ph.D. in Electronics and Telecommunications

    Oct 2024 — Present

    Politecnico di Torino, Torino, Italy. Research areas: semantic communications over optical fiber channels; ML fairness and bias mitigation for optical network QoT estimation; and soft-failure classification.

  2. M.Sc. in Computer & Communications Engineering

    2022 — 2024

    American University of Science and Technology, Beirut, Lebanon. GPA: 3.70 / 4.0.

  3. B.Sc. in Computer & Communications Engineering

    2018 — 2022

    American University of Science and Technology, Beirut, Lebanon. GPA: 3.92 / 4.0 — High Distinction, President's Award.

Publications

  1. Enhancing Reliability of ML-Based Lightpath QoT Estimation via Bias Mitigation Techniques

    Journal of Optical Communications and Networking (JOCN) · Under Review

    A three-stage bias-mitigation pipeline (pre-, in-, and post-processing) cuts group-wise disparity across modulation-format groups by over 90% with accuracy loss within 0.2 points; XAI analysis confirms the mitigated models preserve baseline decision patterns.

  2. Enhancing Reliability of Lightpath QoT Estimation Models using Bias Mitigation Techniques

    International Conference on Transparent Optical Networks (ICTON) 2025 · Barcelona, Spain

    Framed fair lightpath QoT estimation with modulation format as a protected attribute and proposed a three-stage pipeline, improving three fairness metrics by 75–95% while staying within 0.2% of baseline accuracy and ROC-AUC.

  3. Risk-Aware Machine Learning-based Approach for Lightpath QoT Estimation in Optical Networks

    International Conference on Optical Network Design and Modelling (ONDM) 2025 · Munich, Germany

    Proposed a Hybrid Ensemble that allocates model capacity by SNR margin, plus three risk-aware metrics. Achieved up to 7% lower prediction error, 8% lower EDR, and 7% lower RD versus the best baseline on two SDM CONUS-topology datasets.

  4. Comparative Analysis of Liquid Neural Networks and Incremental Learning for Stock Market Prediction

    International Conference on Control, Automation, and Instrumentation (IC2AI) 2025 · Beirut, Lebanon

    Compared Liquid Neural Networks with Incremental Learning for stock-price forecasting; LNNs delivered superior predictive accuracy and robustness on Tesla and Apple datasets.

  5. Predicting Automotive Vehicle Engine Health using MLP & Logistic Regression

    International Conference on Innovations in Computing Research (ICR) 2024 · Athens, Greece

    Developed an engine health prediction model combining MLP for feature extraction and Logistic Regression for classification, enabling early fault diagnosis from vehicle sensor data.

My skills

  • ML & DL

    Deep learning, model training, and explainable AI (XAI).

  • Research skills

    Problem framing, experiment design, and academic writing.

  • Dev tools

    Python, PyTorch, Git, and reproducible ML pipelines.

  • Optical Networking

    Optical fiber channels, QoT estimation, and simulation.

Publications

Contact

Contact Form