• Analyzing and mitigating overfitting vulnerabilities in machine learning models trained on large-scale datasets, ensuring improved robustness and accurate encoding of physical relationships.• Curating, filtering, and partitioning large-scale datasets for training, cross-validation, and testing, enabling experimentation of robustness and reliable model performance.

• Integrating virtual reality into seminars and workshops to connect students in Puerto Rico and Florida.• Developed and taught undergraduate courses in scientific computing, covering data analysis, simulation, computational biology, and scripting in Python and R.• Designed and managed an online classroom platform to guide students through research initiatives and graduate school preparations, fostering a supportive environment.

• Analyzed spectral data from ocean models to forecast rogue wave formations in the deep sea.• Derived asymptotic models of rogue wave distribution, leveraging Python and Mathematica for simulations and data analysis.• Built efficient and stable numerical simulations for studying the distribution of wave amplitudes using numerical methods on Riemann surfaces.

• Collaborated with teams of three to solve complex computational problems using efficient algorithms and data structures in C++, Java, and Python.• Promoted STEM education by authoring problem sets and volunteering as a judge at junior programming tournaments, engaging high school students in programming and algorithmic thinking.

• Collaborated with a team of six in researching and designing a machine-learning model for detecting power grid anomalies.• Developed a graph neural network using PyTorch, Pandas, and NetworkX that reconstructs and analyzes network traffic from the UNSW-NB15 dataset.• Identified and corrected biases in training data, improving model accuracy from around 60% to 94% after retraining.

• Investigated ways to use formal methods to strengthen the reliability of machine-learning models.• Developed a differential calculus for System F, creating a mathematical framework for verifying the correctness of machine-learning training algorithms.• Traveled and presented research at multiple undergraduate symposiums in the U.S., leading to an invitation to present at the 35th International Conference on Information and Intelligent Systems in Croatia.