-

一个非常有用的开源代码

CONTRIBUTE YOUR CODE

Do you have some useful C++ code related to Machine Learning, Data Mining, or Artificial Intelligence? Please consider donating it to the project. Click on the "Help Us" link (on the left). If you're too busy or lazy to really get involved, at least consider sending us the code. Maybe we'll find time to polish it up and integrate it. If you'd like access to our public Subversion repository, just click on "Help Us" and send an email. 



PROJECT STATUS



2006-05-25

Added a new clustering algorithm, a discreet path search algorithm, did a good deal of general code clean up, integrated some changes contributed by Roger Pack into various classes, and fixed about ten bugs. Also thanks to Kevin Kemp for getting it to build on Mac without having to include a special framework in the package. I also added some unpolished tools for making charts and added a link to the API docs from the main menu. 


2006-03-23

This is a bug fix release. If you were getting a build error in GKeyboard.cpp on Windows, that's fixed now. 


2006-03-11

Ported to Mac OSX Tiger. Thanks to Helaman Ferguson for much of this work. It now works on Mac, Linux, and Windows. 


2006-02-25

Added a new efficient neighbor-finding class. The KNN algorithm is much faster now. Fixed many bugs. Added a ray-tracing demo. This isn't really related to machine learning, but I have some plans in the future to try combining it with learning algorithms to do model reconstruction. The manifold learning demo demonstrates both unsupervised and semi-supervised manifold learning now. 


2005-12-31

Added a particle swarm algorithm and some other search algorithms. Fixed some issues with the genetic algorithm. The Neural Net interpolation demo now compares several search algorithms. (Backprop is the clear winner, but it's not a totally fair comparison because backprop is running in incremental mode while the others are doing batch mode.) 


2005-12-17

This is mostly a bug-fixing release. 


2005-12-03

Added automated tests for several of the classes. 


2005-11-27

The project is now released to the public. It builds in VS6 on Windows and with G++ on Linux. Here's some hilights of the initial release... 


Machine Learning Algorithms GNeuralNet A standard sigmoid-based artificial neural net and backpropagation implementation. Polished and well tested 
GNaiveBayes A standard naive bayes classifer. Well tested. 
GDecisionTree A decision tree algorithm based on the C3 algorithm. Real output values are not supported yet. Some automated testing has been done. 
GKNN An implementation of the K-nearest neighbor algorithm. Doesn't yet support pruning or other standard features. Works, but currently uses a slow neighbor-finding algorithm. 
GLLE The Local Linear Embedding manifold learning algorithm. Works. Tested. 
GGenetic A standard genetic learning algorithm. Not very well tested yet. 

Data Mining Tools GArff Parses and abstracts data in the ARFF format. Tested with over 30 ARFF files. 
GStemmer Wraps the Porter Stemming Algorithm Some testing has been done. 
GHttp A simple class for retrieving documents from the web. This class would make writing a web-runner really easy. Reliable enough to use, but some bugs are known to exist. 
GHtml A tool for parsing HTML documents. Use in conjunction with GHttp. Incomplete, but sufficient for most simple purposes. 

Misc Tools
GArray Dynamic arrays. Useful for a million purposes. Polished and thoroughly tested 
GAVLTree A standard AVL tree. Ideal for priority queues. Polished and thoroughly tested 
GBits Some bit-twiddling routings like Gray-code conversion (used in genetic algorithms), etc. Only somewhat tested 
GBitTable In case you need a big array of bits for something Not well tested 
GEZSocket Makes network communication really easy Thoroughly tested 
GFourier The Fourier transform. Known to work 
GFuzzyLogic Some standard fuzzy logic operations. Useful for many AI algorithms. Not well tested 
GHashTable An extremely useful data structure for all sorts of purposes Very well tested 
GImage For image manipulation Pretty stable. Possibly has some bugs 
GMath Some typical math stuff like Gaussians, gamma functions, sigmoids, etc. Somewhat tested 
GMatrix For linear algebra computations. Does eigenvectors. Somewhat tested 
GPolynomial Computes multi-dimensional polynomials using Newton's algorithm Not very well tested 
GPointerQueue Another general purpose and very useful data structure Very well tested 
GTime Useful for benchmarking your algorithms Well tested 
GXML A light-duty XML parser. Somewhat tested