Interests

My interests are not constraint to those who are described in my curricular expedient. In here they are presented some of my interests, hobbies, concerns, worries linked to science and technology. The classification in fields is merely an orientation. The fields and my specific interests are in an overlapping mess.

• Computer Science
  • Data Engineering
  • Artificial Intelligence and Machine Learning
  • Data Science
  • Programming
• Mathematics
  • Optimization
  • Control Theory and Operations Research
  • Information Theory
  • Learning Theory
• Physics
  • Statistical Physics
  • Non-linear dynamics
  • Complex Systems
• Economics and Digital humanities
  • Marketing
  • Finance
  • Macroeconomics
  • Game Theory
  • Digital Humanities

Computer Science

Data Engineering

• Flexible and scalable data architectures for highly parallelization requirements. Using Cloud Computation to run applications able to handle huge data volumes in an efficient way.
• Data pipelines process and new software to deal with multiple heterogeneous sources, automatize transformation and accelerate the process of populating the databases.
• Data flows of integrating data and building generic data models to serve analytical purposes.
• Microservices for the integration of machine learning techniques in the new requirements data infrastructure. The Data science in a IoT world requires an adapted infrastructure to deal with that, not only new machine learning techniques.

Cloud infrastructure

• Create cloud infrastructure for DataManagement, MLOps and automation.

Artificial Intelligence and Machine Learning

• Small data science and One Shot Learners: there are problems with too few samples but high dimensionality. In this problems we can learn the “mechanics” of each sample in order to extract and use that information. It is been developed mainly in the context of OCR problem but it could be very useful for problems in digital humanities.
• The cooperation of logic and statics in new trends of AI. The new machine learning techniques brought us far in the path of seeking true artificial intelligence. But it still seems that we not need only data-driven techniques but a strong cooperation with logic techniques.
• The paper of ecology in the use of evolutionary techniques in ensemble learning.
• Generative Adversarial Networks to make more robust to random noise learners.
• Reinforcement learning and the use of Control Theory to automate decisions.

Data Science

• Spatial Data Analysis and GIS, and its applications to the new urbanism required in the world of IoT and reactive and smart spaces and services.
• Mobility optimization by understanding patterns and using proper statistical matching algorithms.
• Data visualization and dashboards. The improvement of my skills in data visualization and mastering D3.js, Tableau and other Data Visualization tools.
• User behaviour understanding. Collective behaviours and which are the patterns they follow to understand the use.

Programming

• Python development. From pure data science applications to more web-based framework to automatize interaction with databases and show visualizations. Python is my main computation language. Improving from the point of view of software design and coding efficiency is a everyday purpose challenge.
• R programming language. Mainly for exploratory data analysis, at least in the last times.
• Julia programming language. It is a high-level dynamic programming language designed to address the needs of high-performance numerical analysis and it has a fast growing Machine Learning community. It has the possibility to call C and Fortran function through ccall and Python code through PyCall package. It allows concurrent, parallel and distributed computing. I am still in playing level.
• Software engineering. The art of better programming optimal structures for the specific problems. The correct use of SOLID principles, simplicity and correct adaptation of the software patterns to the problems.

Mathematics

Optimization

• PSO methods and the use of the knowledge about dynamical systems to tackle the problems of optimization.
• Genetic programming and discrete optimization. Combinatorial problems are in everyday life. That is why the interest of that type of problems come to me from practical reasons.
• Methods of optimization for multi-peak problems. One of the most amazing and challenging problems you can have in optimization are those which very different solutions can have similar fitness scores. In those problems hill climbing like solvers or traditional solvers cannot do anything. More imaginative and distributed solutions can do good job.

Control Theory and Operations Research

• Optimization of transport systems.
• Optimization of supply-chain management systems.
• Urban planning. As Ildefons Cerdà though, we must use urbanism to optimize and improve social and economical situation of our cities. Specially in the next demographic era of cities, we should be aware of the problem and search the best solutions for that. That problem involves optimization, complexity theory and amazing collection of interdisciplinary fields that make it really sexy problem.

Information Theory

• Information theory and its applications on inferring structural connections in time-series data.

Learning Theory

• Multi-armed bandits problem and its implications with learning theory.

Physics

Statistical Physics

• Applied statistical physics to social sciences.
• Understand renormalization techniques and its implications in symmetry breaking.
• Connections between statistical physics and learning theory.

Non-linear dynamics

• Understanding recurrence and non-linear reactions in social world to proper modeling.
• Study the basis of Artificial neural networks and its limits as the Spiking Neural Networks models.

Complex Systems

• Use of complex networks theoretical framework to better and more intuitive clustering algorithms.
• Study and understanding of fashion processes and contagion processes and its connections with the social networks structure.
• Study how network infraestracture can help prediction of collapse or ensure network growth.

Economics & Digital humanities

Marketing

• Prediction of customer actions and reactive marketing.
• Launching target marketing campaigns through different channels (mail, in-app comunication tools, social media, …) to push customer actions.

Finance

• Inferring structure over networks and find correlations with other probably structure networks (boards of companies, social network, politic-driven or regions-driven networks) in order to help predict and control financial observables.
• Understanding of the bubble dynamics and the collapse of the economics.
• Customer scores which profile customers and help companies to understand their customer basis and better create strategies in a lower granularity level possible.

Macroeconomics

• Creation of artificial economics and self-organization. Study of real economical trends from simulation of microeconomics.
• How complexity study (as the economic complexity index) can explain and help to predict economic growth.
• How statistical physics can help to infer macroeconomic features.
• Political economics and how to control incentives to help to improve certain micro and macro economic variables.

Game Theory

• The study of power in the social ecosystems of representation when there is some topological latent structure.
• Control and balance ecosystem structures (ecology, games development, teams management and multi-agent systems development).

Digital humanities

• Understanding collapse of civilizations and social structures.
• Cultural evolution, post-humanity road-map and its implications.