As a Computer Engineering student at the Polytechnic University of Madrid (UPM), I have gained a comprehensive understanding of computer science and engineering principles. In addition, during my academic exchange year at the Swiss Federal Institute of Technology Lausanne (EPFL), I have enriched my knowledge and skills in the fields of artificial intelligence, computer vision, and natural language processing through courses in the Master's in Data Science program.
In the future, I hope to use my skills and knowledge to make a positive impact on the world by applying computer vision and natural language processing to solve real-world problems. For example, I am interested in using computer vision to improve the accuracy of medical diagnoses, to create safer transportation systems, and to protect wildlife. I am also interested in using natural language processing to improve the efficiency of business operations, to make it easier for people to access information, and to facilitate communication between individuals who speak different languages. I believe that by leveraging the power of technology, we can make a difference and create a better future for all.
I am passionate about learning the theory that is pushing the cutting edge of ML.
I want to give computers the ability to understand text the same way we can.
I enjoy creating models using deep learning algorithms.
I am dedicated to developing valuable and useful software applications.
I love telling a story. Getting to the heart of a problem and coming up with a solution.
Data mining helps organizations extract valuable insights from large data sets.
Take a look at my recent work.
This data story analyzes the representation of female characters in the film industry and examines the impact of the "Smurfette principle," in which media tends to feature a disproportionate number of male characters compared to female characters. The story uses data on film and cast, as well as film summaries, to compare the representation of male and female characters and actors and actresses in terms of popularity and the types of roles they play. It discusses the potential impacts of underrepresentation of women in film on societal attitudes towards women and the importance of improving representation in the industry.
This publication describes the use of a convolutional neural network (CNN) called U-Net to segment roads in satellite images. We preprocess the data by applying data augmentation techniques and dividing the images into smaller patches, which are then used to train the U-Net model. The model is evaluated using various metrics, including accuracy and F1 score, and is found to achieve an accuracy of 94.5% on the test set. We also perform an error analysis to identify areas for improvement. The results suggest that the U-Net model is a promising approach for this task.
This publication describes a machine learning model based on regularized logistic regression that was developed to perform binary classification for the detection of Higgs boson production at CERN. The model was trained on a dataset consisting of 250,000 samples, each described by 30 features, and achieved an accuracy of 79%. My teammates and I also outlined a process for handling missing data, removing correlated features, and standardizing the remaining features in order to improve model performance.
This project involves implementing the A* algorithm to find the optimal path between two metro stations in the Kyiv metro system, based on both the distance and the number of transfers required. The project was implemented using NodeJS and includes an interactive user interface that allows the user to select the two stations they want to calculate the optimal route for.
