Embedded Files
My current academic research interests include Statistics, Data Science, Computational Sciences, and Applied Mathematics, with a focus on Machine Learning, particularly in medical and environmental science applications. My Ph.D. dissertation focuses on using Machine Learning to predict and model the temporal variability in CO₂, CH₄, and N₂O fluxes from human-made aquatic systems such as hydropower and fishponds using global geospatial data, where capturing temporal patterns is essential for understanding ecosystem behavior. Some of the research techniques I utilize are Random Forest, Bayesian Hierarchical Models, Neural Networks, RNN’s, Meta-Learning, Transformers and Conditional Variational Autoencoders as GenerativeAI. These can be used to target high emitting reservoirs for mitigative management as well as to inform the placement of new constructed aquatic ecosystems at the global scale.
Another line of my research focuses on analyzing medical data to estimate the variance of the estimators after using cross-validation and imputation techniques in classification problems. I was awarded the Best Poster Award (3rd place) at the Student Poster Competition at the Conference on Statistical Practice and an Honorable Mention at the Mathematical Congress of the Americas for this work, presented in my project “Theory and Application of ROC Curves with Cross-Validation Estimators for Clinical Data with Missing Observations.” This project was particularly interesting to me because I explored theoretical mathematics methods, as well as the applications using real-world data and simulations to focus on techniques associated with ROC Curves, and the variance of the estimators obtained through cross-validation. The main goal in this project was to describe the mathematical foundations of imputation techniques, formalizing the theory in which they are based and how these methods are applied to health data; and then understand how the applications are related to the mathematical methods we use.
Even though my background as an undergraduate was pure Mathematics, during my PhD. program, I decided to transition to Applied Mathematics and Statistics. In my current research, I design and apply Machine Learning and AI methods for datasets that are highly variable, sparse, or contain missing values.
Research Experience
Graduate Research Assistant at The Environmental Change and Sustainability (ECS) Lab
Developed predictive models, including Random Forest and Lasso regression, to estimate greenhouse gas fluxes from reservoirs, utilizing global geospatial data and performing feature selection.
Modeled the temporal variability of greenhouse gas (GHG) emissions, specifically carbon dioxide (CO2), and methane (CH4), from freshwater aquaculture systems in Brazil using techniques such (RNN, LSTM, GANs, transformers).
Developing convolutional neural network (CNN) models to classify fish ponds in the Minas Gerais state of Brazil using satellite imagery.
Fall 2023-Outgoing, University of Texas Rio Grande Valley
Research Assistant Aquatic Ecology Lab (LEA)
I had the incredible opportunity to travel to Brazil for fieldwork and do research on greenhouse gas emissions at the Ecology Lab of the Federal University of Juiz de Fora. While my background is in Mathematics, for the past year, I’ve been applying machine learning techniques to model greenhouse gas emissions in human-made aquatic ecosystems.
One of the biggest challenges I’ve faced is dealing with the complexity of this type of data, particularly because collecting it is both time-consuming and expensive and the variability in the data is huge. During my time in Brazil, I finally had the chance to see how this data is collected. I worked on fish ponds, taking daily measurements. It was an amazing experience because I can realized why working with this data is so challenging.
This experinece not only helped me to understand better my research and think about new methods that I can apply, but also taught me how to take measurements of greenhouse gas emissions and the factors associated to these emissions. I learnt how to use a chamber, and identify the key contributors to greenhouse gas emissions in fish ponds.
I am grateful for this opportunity, which allowed me to learn about a field that I had never imagined I would delve into, which I found amazing and interesting, And also allowed me with the opportunity to deal with challenges where I can merge my background in Mathematics and Data Science with applications to real world problems.
Summer 2024, Federal University of Juiz de Fora, Brazil
Research Group TREA-Techniques for Representations of Algebraic Structures
As an undergraduate student at the National University of Colombia, I got the opportunity to study Representation theory of Algebras, particularly, Representation Theory of Posset. All of this started in 2017 when Prof. Gonzalo Medina invited me to join his research group as part of a program of the National University. Therefore, I became an active member of the Research Group and began research on Representation Theory of Posset and algebras. First, we research Posset Representation’s Differentiation Techniques such as differentiation concerning a pair of points, and differentiation with a maximal point. Second, we researched Frobenius Algebras. Afterward, Prof Gonzalo Medina oversaw an important project about relationships between Representation Theories of Algebras, the Theory of Knots, and Combinatorics with some applications, and I got the opportunity to be a part of it. On this project, we researched and worked together on the four subspace problems.
Fall 2017-Fall 2020, National University of Colombia
The research and the project were the greatest contributions to my academic life. I feel this because it enabled me to broaden my knowledge about the topics discussed, learn how to conduct research in Mathematics, develop my communicative abilities, and realized that I wanted to teach and share my knowledge with others. Furthermore, it allowed me to participate in national and international academic events where I disseminated the results of the investigation. CIMPA Schools, the National Congress of Mathematics in Colombia, courses and conferences were among the activities in which I was involved.
Participating in these events allowed me to obtain a deeper understanding of the topics I had previously studied, as well as meet people from all over the world who had extensive mathematical expertise. As a result, I learnt about Nakayama’s Algebras, Artin Allegra’s, Commutative Algebra, Tensor Categories, Hopf Algebras, Quantum Theory, and Representation Theory of Algebras when I attended the CIMPA school of Geometric and Homological Methods in the Representation Theory of Associative Algebras and Their Applications in Medellin/Colombia and CIMPA school of Algebras and Tensor Categories in Argentina. In addition, these experiences helped me to understand the importance of mathematical communities where the knowledge is disclosed and different cultures and ways of thinking are exposed. Definitely, participating in events and programs like these was an exceptional academic and cultural experience.
Page updated
Google Sites
Report abuse