Mohamad Mansour

Machine learning researcher & engineer

Mohamad Mansour

I work on explainable AI, reinforcement learning, NLP and computer vision.

Get in touch Selected work

Now at Whish Money, working on internal AI projects.

Selected work

  1. Present Project Whish Money

    Internal AI projects

    Internal, so there’s nothing to link to.

  2. Paper American University of Beirut

    CEnt: An Entropy-based Model-agnostic Explainability Framework to Contrast Classifiers’ Decisions

    Explains a classifier’s decision by generating feasible feature tweaks that would change it.

    Keywords: deep learning, explainable AI, contrastive AI, counterfactual prediction, PyTorch, Python

    Read the paper Code on GitHub

    The CEnt pipeline in five steps: train a VAE on the full dataset; sample from the VAE encoding and train a decision tree on the original representation; build a graph with the edge construction technique; locate the factual and search for the closest counterfactual node; generate counterfactuals with random perturbations.
    Fig. 1 The CEnt pipeline, from VAE to counterfactuals.
    Handwritten digits 5 and 6 shown as the original, as a CEM counterfactual, with CEM visual contrast, and with CEnt visual contrast.
    Fig. 2 Visual contrast on handwritten digits: CEnt against CEM.

  3. Paper American University of Beirut

    On the Evaluation of the Plausibility and Faithfulness of Sentiment Analysis Explanations

    Metrics for judging the explanations of sentiment analysis models: whether the extracted “rationales” faithfully explain the prediction, and how well explainability methods agree with human judgment on a home-grown dataset.

    Keywords: deep learning, explainable AI, faithfulness, PyTorch, Python

    Read the paper on Springer

    Evaluation framework: a sentiment analysis model and an explainability model produce rationales for a sentence. Faithfulness is measured as accuracy on the explanations; plausibility is measured against ground-truth rationales using precision, recall and fallout.
    Fig. 3 Evaluating explanations for faithfulness and plausibility.

  4. Research American University of Beirut CYENS

    Navigating Agent

    Room navigation for embodied agents in Habitat, comparing combinations of reinforcement learning algorithms and training methods.

    Keywords: reinforcement learning, computer vision, metacognition, PyTorch, Python, multiprocessing

    About Habitat

    The Habitat point-goal navigation task, as illustrated on aihabitat.org.
    Fig. 4 The Habitat point-goal navigation task, via aihabitat.org.

  5. Project

    Course Interactive Website

    During Covid-19, finding one specific lecture recording meant a lot of searching. I built this site so my classmates and I could find things faster.

    Keywords: interactive website design, HTML, CSS, JavaScript

    Visit the site

    Screenshot of the HIST 273 course website home page, with a navigation bar and a grid of lecture tiles.
    Fig. 5 The course site’s home page.

  6. Paper American University of Beirut

    Explainable Model for EEG Seizure Detection

    An explainable model for detecting seizures from EEG, to help diagnose epilepsy. It also ranks the features that trigger a prediction.

    Keywords: deep learning, explainable AI, feature extraction, TensorFlow, Python

    Read the paper

    EEG recordings are turned into connectivity features and fed to a deep network for seizure prediction. The model weights are then evaluated to rank the features that trigger a seizure prediction.
    Fig. 6 From connectivity features to ranked seizure triggers.

  7. Project American University of Beirut Asurion

    Extracting Insights

    Classifying reviews and extracting insights from them, using unsupervised data augmentation.

    Keywords: natural language processing, unsupervised data augmentation, Python, Keras, scikit-learn, NLTK

    A text preprocessing pipeline (remove non-alphabetic characters, lowercase, lemmatize, n-grams) above a diagram of unsupervised data augmentation with supervised and unsupervised consistency losses.
    Fig. 7 Review preprocessing and unsupervised data augmentation.

Experience

Contact

moemansour03@gmail.com