## Projects

**Thesis Projects**

- Machine Learning Methods:
- We are looking at new algorithms for
- Kernel Learning
- Metrics for new clustering algorithms
- New methods for deep learning
- Large Scale Optimization for Machine Learning
- Here we are looking at large scale optimization methods for machine learning applications where the input is exponential in nature. This has applications in the areas of Data Mining and Big Data.

- We are looking at new algorithms for
- Artificial Intelligence
- We are looking at new approximation algorithms for learning plans to
- Minimize space complexity of the state representation
- Improve planning trough Reinforcement Learning

- We are looking at new approximation algorithms for learning plans to
- Social Networks
- Discovering influences in social network represented by a graph with thousands of nodes. We are looking at light algorithms to handle Big Data sets which are able to discover possible thrends in social networks. This has applications in:
- Web Search
- E-commerce
- Page Ranking

- Discovering influences in social network represented by a graph with thousands of nodes. We are looking at light algorithms to handle Big Data sets which are able to discover possible thrends in social networks. This has applications in:
- Bio-informatics
- Development of new tools using Self-Organization Maps, Bayesian Networks and Graph Methods for looking at cellular cycles, classification of gens, etc.
- Graph representation of cellular cycles
- Simulation of cellular cycles through graph theory

- Fuzzy Logic Applications
- We are looking at several possible applications:
- Modeling of complex rule systems for control of virtual characters for animation.
- Fuzzy Rule Systems for Render in Computer Graphics
- Energy consumption froecasting
- Learning of Fuzzy Systems

- We are looking at several possible applications:
- Computer Vision
- Hyperespectral Imagenery: We are looking an more reliable techniques to find endmembers and use them to be able to classify pixels in a hyperespectral image. This has many industrial applications from crop survillance to security systems.
- Face Recognition - We ae looking for algorithms for face recognition of faces in crowds. This has applications in Biometric. Possible algorithms to be used are:
- Fuzzy Pattern Theory.
- Deep Learning.
- Statistical Image Analysis.

- Feeling Recognition - We want to develop algorithms for recognition of feelings for profiling.
- Robust and fast Algorithms for object recognition with applications in massive datasets. This can have applications in:
- Image Web Search
- Biometric
- Medical Imaging

- Advanced data structures for:
- Data Mining. For example, for graph representations or locality sensitive searches under concurrency and paralleism.
- Databases. For example, for fast index access using B+-trees or Skip List under concurrency and parallelism.
- For acceleration of numerical operations like sparse matrix multiplications under concurrency and parallelism.

Página anterior: Publications

Página siguiente: Slide Share