Extended Competent Genetic Algorithm (ECGA) and Philo-Genetic Algorithm (PhGA)
In this Scientific Initiation Project a competent Genetic Algorithms denominated ECGA (Extended Compact Genetic Algorithm) was implemented. This algorithm is more reliable, efficient and accurate than common Genetic Algorithms, being capable of solving large scale problems by creating a probabilistic model of the population. This is currently the state of art in the modern artificial intelligence field. As an extension to the project, improvements in the ECGA were tested. At the end, by combining phylogenetic ideas into some parts of the genetic algorithm and using a different approach to the dynamic of populations throughout the generations, we created a new kind of genetic algorithm with the name of Philo-Genetic Algorithm (PhGA).
We are now writing some papers and finishing the tests, as soon as possible I will post here either a tutorial or a paper about the PhGA.
This project was approved and supported by the FAPESP Institution as a scientific initiation project, therefore I was awarded a scholarship.
Period: January 2009 - Present
Coordinator: Alexandre Claudio Botazzo Delbem
Language: C++
Scholarship: FAPESP Scholarship for Scientific Initiation
In this Scientific Initiation Project a competent Genetic Algorithms denominated ECGA (Extended Compact Genetic Algorithm) was implemented. This algorithm is more reliable, efficient and accurate than common Genetic Algorithms, being capable of solving large scale problems by creating a probabilistic model of the population. This is currently the state of art in the modern artificial intelligence field. As an extension to the project, improvements in the ECGA were tested. At the end, by combining phylogenetic ideas into some parts of the genetic algorithm and using a different approach to the dynamic of populations throughout the generations, we created a new kind of genetic algorithm with the name of Philo-Genetic Algorithm (PhGA).
We are now writing some papers and finishing the tests, as soon as possible I will post here either a tutorial or a paper about the PhGA.
This project was approved and supported by the FAPESP Institution as a scientific initiation project, therefore I was awarded a scholarship.
Period: January 2009 - Present
Coordinator: Alexandre Claudio Botazzo Delbem
Language: C++
Scholarship: FAPESP Scholarship for Scientific Initiation
3D Game Engine - Strategy/Action
As the team leader, I guide the group through the process of creating a complex software engine as well as envision ways to work with artists efficiently.
We are building a 3D Game Engine together with a featured game to shown the capabilities of this engine. The main focus of the Engine is building Fighting, Action, Strategy and Action RPG games.
Since the Engine is still in development little could be said about the game, but it will have a strong focus on gameplay.
Every implementation done in this project is well aware of the necessity of building flexible, extensible and fast code. In this way lot of effort was put to share data in memory among other optimizations and designs.
To make clear my contributions to the project as a programmer, note that everything presented on the video below was programmed by myself, setting aside the texture, models and animations.
Features (implemented by myself):
Libraries: SDL, GLU, OpenGL, OpenAL, Lua, tolua, Boost, Vorbis and Alut.
Language: C++, Lua
State: Active
As the team leader, I guide the group through the process of creating a complex software engine as well as envision ways to work with artists efficiently.
We are building a 3D Game Engine together with a featured game to shown the capabilities of this engine. The main focus of the Engine is building Fighting, Action, Strategy and Action RPG games.
Since the Engine is still in development little could be said about the game, but it will have a strong focus on gameplay.
Every implementation done in this project is well aware of the necessity of building flexible, extensible and fast code. In this way lot of effort was put to share data in memory among other optimizations and designs.
To make clear my contributions to the project as a programmer, note that everything presented on the video below was programmed by myself, setting aside the texture, models and animations.
Features (implemented by myself):
- Animated Characters and static Objects
- Sound Engine (Format support: Ogg, Wav)
- 3D Engine (supports for selection, 3D movement, Action System using animations of skinned characters and list of actions to create combo and to store multiple commands, Support for 3D Model Formats: Md5mesh and .OBJ, among other things...)
- User Modifiable Interface
- Lua Script Support to load and unload Characters, Objects and to define the Scenario. (So artists could easily test different Characters, Objects and Scenarios Sets without the need of compiling the source code)
- VBOs, World specific drawing optimizations, Memory Sharing support for Wav, Ogg, Md5mesh and Obj, among other optimizations.
- Ported to Linux. It was ported for Debian and Gentoo distributions mainly.
- Effect's System featuring a extensible set of effects that are configured via Lua scripts, but that are implemented and extensible in C/C++. In a way that artists have more possibilities and a faster way of testing. And that programmers could set things up for performance.
- Well Documented code in our wiki for the developers and a documentation of how to use and modify the Lua scripts and other settings for the artists who are going to work with scenario design and modeling.
- Basic Artificial Intelligence
Libraries: SDL, GLU, OpenGL, OpenAL, Lua, tolua, Boost, Vorbis and Alut.
Language: C++, Lua
State: Active