Welcome to my GitHub profile! Here, you'll find projects I've worked on, focused on Data Science, Machine Learning, and their applications in various areas such as urban inequality and health.
🎓 Bootcamp in Data Science & AI.
📊 Passionate about solving complex problems through Machine Learning techniques.
🎶 Fields of my projects: LLMs for Education, predictions in Healthcare, and analysis on Transportation and Inequality.
vihsible is a chatbot made as a final team project for the Spanish LGTBI+ Federation. We used Langchain with Cohere LLM in several prompts, that sent information between them in order to know which prompt is to answer the user.
DSTrainer, another LLM project that uses Mistral to help users train on technical Data Science questions. The model evaluates the answer of the user, gives a numerical grade, and makes a follow-up question, depending on how correct the answer of the user was.
AI.lhz: A prediction tool for doctors: an app I developed consisting on two classification models: a GradientBoostClassifier (accuracy of 95%), and a Convolutional Neural Network (accuracy of 87%), which detects brains with or without Alzheimer, and it's degree of advancement.
Check out the app from this link
Transport and Income: Inequality in Madrid: a deep Exploratory Data Analysis (EDA) on the patterns of mobilization in Madrid, which also finds a link between income and the preferred method of transportation of workers within the region.
Some of my code used for Kaggle competitions:
- 🏆 Image Emotions: in this competition the goal was to differentiate the emotions on images, from a range of up to seven emotions. Convolutional Neural Network that reached an accuracy of 57%.
- 🏆 Laptop Regression: the goal was to predict the price of a computer. I used webscrapping (Selenium) to get additional data, and the winner model (GradientBoost with polinomical features) reached a MAE of 181.2.
Languages: Python, SQL
LLMs: Mistral, Cohere.
Machine Learning / IA: Scikit-Learn, TensorFlow, PyTorch, LangChain
Data Analysis: Pandas, NumPy
Visualization: Matplotlib, Seaborn
Deployment: FastAPI, Streamlit, Docker, AWS.
If you have any questions about my projects or want to collaborate on a topic related to machine learning and data science, feel free to reach out: