Current Flows in Electrical Networks for Error-Tolerant Graph Matching

Current Flows in Electrical Networks for Fuzzy Social Network Analysis (FSNA)

Quantum Computers: past, present, and future

Master's Thesis - Current Flows in Electrical Networks for Error-Tolerant Graph Matching

Undergraduate Thesis - X-RIB - 3D modeling tool implementing Pixar's Renderman Interface Bytestream specification using BMRT (Blue Moon Rendering Tools) [Spanish]


X-RIB - Undergraduate Thesis [C++]
[+] source code  [+] documentation

Description: X-RIB was a CAD tool implemented in C++ for both Linux and Windows environments using the QT C++ libraries from Trolltech. The software implemented Pixar's Renderman Interface Bytestream to generate a .RIB file to be rendered using the Blue Moon Rendering Tools. Among its different modules, X-RIB featured a full animation mode with geometry interpolation between frames.

X-Rib - 3D-CAD tool for Linux and Windows

Graph Matching - Master's Thesis [C++]
[+] source code

Description: The thesis presented a novel approach to error-tolerant graph matching based on current flow analysis. Analysis of current flows in an electrical network is a technique that uses the voltages and currents obtained through nodal analysis of a graph represented as an electrical circuit. An algorithm to calculate a similarity measure between two graphs based on the current flows along geodesics of the same degree was proposed. This similarity measure can be applied over large graph datasets, making it feasible to compare datasets of significant size in a reasonable amount of time. The thesis investigated the classification potential of several data mining algorithms based on the information extracted from a graph dataset and represented as current flow vectors. The operational prototype was evaluated on the NCI-HIV dataset.

Delta Rule- Neural Networks [C++]
[+] source code

Description: This program implements the delta rule used in some neural networks for learning. This was a small assignment from my Machine Learning class.

F1 expert system [CLIPS]
[+] source code  [+] documentation

Description: The F1 expert system was designed as part of my Artificial Intelligence class. The goal was to design a sample system that would assist F1 teams with the correct selection of tires, fuel loads, and racecar settings such as front and rear downforce levels. The F1 expert system was created using CLIPS (Quicksilver Beta 12/31/07). The F1 expert system is designed based on the 2008 Formula 1 season rules and tracks. With 18 races across the world, the Formula 1 2008 season presents different challenges to the teams based on different tracks, racecar setups, weather conditions, and driver abilities and styles.

F1 Expert System in CLIPS

Simple learning - Tic-Tac-Toe [Vb.Net]
[+] source code

Description: Tic-Tac-Toe was a program I wrote as part of an assignment of my Machine Learning class where we were using a very simplistic algorithm to try to calculate the weights of a linear learning function as the game progresses. The program featured a virtual player that allowed to play many games in a short amount of time with the purpose of training the program. The linear function was not enough to successfully learn and always win or tie the game but it represents a good learning example.

"Intelligent" Tic-Tac-Toe

Legal Lead Pro [C#]
[+] proprietary code

Description: Legal Lead Pro was a personal project I did with the goal of commercializing it. It featured an advanced search mechanism of arrest records from several counties in the State of Florida. The records were parsed from multiple sources (PDF, HTML, Text) using regular expressions in PHP and imported into a MySQL database that was accessed by the program using web services. The software allowed the user to filter records based on complex criteria and to print different label types, envelops, and perform mail mergers.

Legal Lead Pro - Automatic Lead Generator

Design by Alejandro Gutierrez