👋 About Me
Hi! I'm César Aguirre, a Physicist passionate about Machine Learning, Deep Learning, and Artificial Intelligence, currently pursuing a Master’s in Statistical Computing at CIMAT Monterrey .
🎓 M.Sc. Statistical Computing (2024–2026) – CIMAT, Monterrey
🎓 B.Sc. Physics (2017–2022) – UNAM, Mexico City
My interests lie in Statistical Inference, Optimization, NLP, Computer Vision, Multimodal Learning, and AI applications.
🛠️ Languages & Tools
- 👨🏫 Teaching Assistant at CIMAT (Jan 2026 – Present).
- Courses: Data Science & Programming and Algorithm Analysis.
- 🎤 Selected Speaker – 58th National Congress of the SMM (Oct 2025).
- Presented: "Language Models and Deep Learning for Sentiment Analysis in Spanish".
- 🔬 Research Assistant at CIMAT – Data reduction & clustering.
- 🏅 Honorable Mention – Rest-Mex 2025 (International Rest-Mex Forum)
- Sentiment Analysis on Magical Towns (ranked 7th/70 teams).
- 📊 Outstanding Participation – DSC Datathon 2025 (Tecnológico de Monterrey)
- Innovative applications of data science in industry challenges.
- 🧭 Workshop Speaker – 1st International Congress on Artificial Intelligence (UNAM, 2025)
- Astronomy and Deep Learning: Identifying extragalactic objects.
-
🎵 Hybrid Two-Tower Architecture for Multimodal Music Recommendation (2025) - Developed a SOTA Late Fusion system using SASRec for sequential modeling and a multimodal Item Tower.
- Integrated Audio (ResNet-18), Lyrics (mDeBERTa + LoRA), Vision, and Tabular metadata using InfoNCE loss.
- Managed MLOps lifecycle with DVC and uv for robust reproducibility.
- Tools: Python, PyTorch, Transformers, Computer Vision, MLOps.
-
📈 Stochastic Modeling of Demand in Video Games (Steam) (2025) - Developed a SARIMAX dynamic regression model to predict Concurrent Users (CCU) on the Steam platform.
- Integrated seasonal component analysis and exogenous shocks (sales/updates) to quantify "Download Lag" in AAA titles.
- Optimized infrastructure resource management and reduced operational risks.
- Tools: Python, Time Series, Statistical Modeling, Forecasting.
-
⚽ Multivariate Analysis of Premier League Performance (2025)
- Applied factor analysis & MDS to uncover hidden patterns in football analytics.
- Tools: R, Python, Multivariate Statistics.
-
🤖 Hybrid NLP Models for REST-MEX 2025 (2025)
- Designed a hybrid system combining Transformers (BETO) + Word2Vec + MLP.
- Tasks: sentiment analysis, destination classification, magical town recognition.
- Results: Macro F1 = 0.60 (polarity), 94.6% accuracy (destination).
-
🪐 Neural Networks for Extragalactic Object Classification (Thesis, 2024)
- Used FITS astronomical data + CNNs to classify galaxies and quasars.
- Tools: Python, TensorFlow-Keras, SQL.
-
🌟 Box-Cox Transform for Stellar Mass–Luminosity Relation (2024)
- Applied Box-Cox transformation to improve prediction of stellar luminosity.
- Tools: R, Statistical Inference, Astronomy.
- ✉️ Email: cmac.medios97@gmail.com
- 💼 LinkedIn: linkedin.com/in/cesar-ac
- 📄 Curriculum Vitae (PDF)
“Our future is a race between the growing power of our technology and the wisdom with which we use it. Let’s make sure that wisdom wins.”
— Stephen Hawking

