AcStudio is a product for determinate the type rule (chaotic, stable, decreasing, increasing, chaotic - complex )

=OpenCapture= OpenCapture is a java open-source document capture system. Our goal is to provide a capture system that can cross platforms. In the current market all major document capture systems are designed to run on the windows platform. We hope to provide a good alternative. == Update ...

=What is !FastRandomForest?= !FastRandomForest is a re-implementation of the [http://www.stat.berkeley.edu/~breiman/RandomForests/cc_home.htm Random Forest] classifier (RF) for the [http://www.cs.waikato.ac.nz/ml/weka/ Weka] environment that brings speed and memory use improvements over the origi...

CARS, currently coupled with partial least squares linear discriminant analysis(PLSLDA), is an efficient strategy for identiying an optimal subset of variales for classification. An exponentially decreasing function(EDF) is introduced into CARS for highly efficient screening of variables, which is e...

==FAQ== ===1. Which Algorithms are implemented in Snabler?=== (So far) A Parallel Machine Learning Classifier for Hadoop Streaming ===2. Why the name Snabler?=== The word Snabler is the Danish and Norwegian plural of an Elephant's Trunk (e.g. the Hadoop elephant), and shapewise referring...

==YARA in a nutshell== YARA is a tool aimed at helping malware researchers to identify and classify malware samples. With YARA you can create descriptions of malware families based on textual or binary patterns contained on samples of those families. Each description consists of a set of stri...

=Randomized Decision Trees= Ertree are more or less extremely randomized decision trees implemented in python. It lets you choose how randomized the training will be and which methods to use. ==Flavor== {{{ $ ../ert.py -t train.jf.gz -T test.jf.gz -e 5 -s ent --leafsize 5 -c 50 --probs tes...

A collection of classes to find the relevance of candidate terms from a text corpus to any particular topic/sub-domain using a term classification driven approach. The system utilizes the lexical and contextual profiles of the candidate and domain-representing "Resource Terms" (Seed and Ontological)...

Traditional Bag-of-Words (BoW) model is a two-layer representation. The unsupervised clustering is applied first to generate the BoW histogram and the maximum margin classifier e.g. SVM is then used to classify data based on BoW histogram. This project designs a learning algorithm to learn the maxim...