Machine Learning, Deep Learning, Web Desing, Natural Language Processing, Data Science. Python, JavaScritp, HTML&CSS
Machine Learning, Deep Learning, Web Desing, Natural Language Processing, Data Science. Python, JavaScritp, HTML&CSS
I recently concluded my Master's Degree at the University of Ottawa in Computer Science with a Concentration in Artificial Intelligence (GPA 3.91/4). My graduation project focuses on CyberSecurity, specifically how to integrate Large Language Models with traditional Machine Learning algorithms to analyze both system logs and raw IP packets to improve the prediction of malicious cyberattacks. I studied for a bachelor's Degree in Electronics and Telecommunications. A couple of years ago, I had the opportunity to work with a multinational Financial Institution in the technology sector for four years. Please scroll down to find out more about my projects and related interests.
Interests
Take a glance at some of the topics I have been working on
interest 1
As my Master's final project, I developed an ensemble model which consists of a DistilBERT / BERT pre-trained and fine-tuned model to extract relevant features from system logs. Together with a Random Forest Binary Classifier trained to analyze raw IP packets and detect DNS attacks. Resulting in a hard voting algorithm which improves the prediction of possible Denial of Service attacks.
More infointerest 2
Experience training and deploying models based on neural networks such as MultiLayered Perceptrons, Convolutional Nural Networks, Recurrent Neural Networks, Transformer-Based Models such as BERT and Generative Adversarial Networks using standard libraries such as Pytorch and Tenforflow.
DL Projectsinterest 3
Experience using Recurrent Neural Networks and BERT-based models using frameworks such as Tensorflow, Keras and HuggingFace. Experience with NLTK library to manually extract features from text and create models based on RNN or BoW/Word2Vec.
NLP Projectsinterest 4
Experience training and configuring different types of Convolutional Neural Networks, together with Generative Adversarial Networks. The research done in this field consisted of training different kinds of GANs to translate a sketch of an image into an Impressionist painting. This is to have an alternative to traditional Data Augmentation methods to improve the quality of image generation.
Img Procc Projectsinterest 5
Desing and Analysis of User Interfaces using HTML and CSS. Complementing these skills with Bootstrap and React. Also some basic knowledge on JavaScript and NodeJS to create endpoints and APIs.
HTML JS Projects