Hi, my name is Francisco Valerio! I am 27 years old and I am a programmer and a master in Physics from México 🇲🇽. As a physics & math enthusiast with a strong analytical background (specially in complex systems, stochastic processes, and artificial intelligence models) I'm passionate about applying my skills to solve complex problems using advanced data science and machine learning techniques.

• 📊 Data analysis: I have proficency in statistical analysis, data manipulation and abilities to visualize data insights.
• 🤖 Machine Learning: Solid and expertise in designing and applying machine learning models, specially Artificial Neural Networks, including Physics Informed Neural Networks (PINN), Convolutional Neural Networks (CNN), Recurrent Neural Networks (RNN) and Generative Adversarial Networks (GAN). Also, I have experience with regression models (Random Forest, Linear Regression & PCA) aswell with classification models (SVM, Logistic Regression & K-Neighbors).
• 🧑‍🔬 Scientific computing: I've developed a robust skill set centered around algorithmic solutions for complex mathematical problems. My experience encompasses techniques such as numerical differentiation and integration, Fourier analysis and the solution for high dimensional non-linear partial differential equations.
• 🙋‍♂️ Adaptability and team work: I excel in adapting to new challenges, learning quickly and applying my knowledge to achieve goals. By collaborating with teams, I think I contribute to a synergetic effor that magnifies and optimizes the results, believing that collective intelligence surpasses individual efforts.

• Image classification
• Price recommendation system
• Marketing campaigns success predictor (Winner project for the BEDU Santander Prototype Day 2022)
• Stochastic optimal control in epidemiology
• Churn detector

Right now, I am working as an external senior advisor for the Instituto Electoral de Estado de Puebla, in the statistics and data science department.

Computer vision, Cloud computing & Time Series Analysis

💼 Linkedin

📧 E-mail

According to Yann Le-Cun:

• LLM: 1E13 tokens x 2 bytes/token = 1E13 bytes.
• 4 year old child: 16k wake hours x 3600 s/hour x 1E6 optical nerve fibers x 2 eyes x 10 bytes/s = 1E15 bytes. In 4 years, a child has seen 50 times more data than the biggest LLMs. Impressive, right? 😆