irene
-
- Username:
- irene_e
- Tags:
- none
Networking
No social network accounts found.
irene_e's Projects
-
-
- Members:
- 2
Environment where you can run genetic algorithm and observe results. School project.
Observer
-
- Members:
- 4
JANE is an Artificial Neural Networks library. It supports currently Perceptron, Adaline, Multilayer Perceptron, Competitive Network architectures; and it has experimental support for Kohonen SOFMs. JANE is part of my undergraduate thesis project. The source have just been extensibly refactored so there should be some bugs to squash.
Observer
-
- Members:
- 10
JScience-JPA is a framework built on top of JScience and Hibernate that facilitates persistance of JScience units to a database. JScience handles both explicit unit conversions and the implicit conversions entrained by mathematical operations between two or more unit values. The JScience-JPA library maps these unit values as custom types at the ORM configuration level, allowing a system to persist these unit value objects as though they were types natively supported by the RDBMS.
Observer
-
-
- Members:
- 4
NAALS is an attempt to evolve new ways of walk for 4 legged creatures. Main application is being developed in JAVA. Main program will use specialized Genetic Algorithm with floating point representation of genes. Genetic operators and structure of genes will be designed to provide a lot of freedom to main GA. Simulation environment is developed in C++ using two 3rd party engines - nVidia PhysX and Irrlicht. The project is now in very early state of developement.
Observer
-
- Members:
- 4
The open framework for analysis, exchange and integration of neuronal data. At this moment project is under construction.
Observer
-
- Members:
- 4
System, that should recognize russian speech, NOT command system. It is based on ear modeling by the instrumentality of neural network. At this moment sound analysis and neural network are partly developed.
Observer
-
-
- Members:
- 12
Since there are many java projects that rely on AI, this project aims at constructing a supervised neural network framework (multi layer perception model), main objective being ease of usage. What makes this project different from others is that it provides a really easy to use neural net creation, training API's along with appropriate customization to support newer training techniques.
Observer
