Welcome to OpenDEN
Custom Search
Artificial Intelligence
Artificial Intelligence :
The use of programs to enable machines to perform tasks which humans perform using their intelligence. Early AI avoided human pychological models, but this orientation has been altered by the development of connectionism, which is based on theories of how the brain works. In connectionism, complex functions, including learning, involve the transmission of information along pathways formed among large arrays of simple elements
From: www.filosofia.net/materiales/rec/glosaen.htm
TiMBL
Platform : Unix like
License : Academic research project. Contact for licensing terms.
Appears in these categories : Artificial Intelligence
License : Academic research project. Contact for licensing terms.
Appears in these categories : Artificial Intelligence
TiMBL is a program implementing several Memory-Based Learning techniques. TiMBL stores a representation of the training set explicitly in memory, and classifies new cases by extrapolation from the most similar stored cases.
On top of the classic k-NN classification kernel, several metrics, algorithms, and extra functions are implemented in TiMBL:
Feature weighting for dealing with features of differing importance: information gain, gain ratio, chi-squared, shared variance
Stanfill & Waltz / Cost & Salzberg modified value difference metric (MVDM) and Jeffrey Divergence (new metric; a symmetric variant of Kullback-Leibner divergence) for making graded guesses of the match between symbolic values
Class voting in the k-NN kernel according to distance of nearest neighbors (inversed, inverse linear, exponentially decayed)
Optimizations for fast classification: conversion of case base to tree
Even faster classification using heuristic approximations: the IGTREE decision tree algorithm and the TRIBL and TRIBL2 hybrids
Emulation of Aha, Kibler and Albert's IB2 algorithm for incremental edited instance-based learning
Exemplar weighting
Fast leave-one-out testing and internal n-fold cross-validation
Multi-threaded server functionality
On top of the classic k-NN classification kernel, several metrics, algorithms, and extra functions are implemented in TiMBL:
Feature weighting for dealing with features of differing importance: information gain, gain ratio, chi-squared, shared variance
Stanfill & Waltz / Cost & Salzberg modified value difference metric (MVDM) and Jeffrey Divergence (new metric; a symmetric variant of Kullback-Leibner divergence) for making graded guesses of the match between symbolic values
Class voting in the k-NN kernel according to distance of nearest neighbors (inversed, inverse linear, exponentially decayed)
Optimizations for fast classification: conversion of case base to tree
Even faster classification using heuristic approximations: the IGTREE decision tree algorithm and the TRIBL and TRIBL2 hybrids
Emulation of Aha, Kibler and Albert's IB2 algorithm for incremental edited instance-based learning
Exemplar weighting
Fast leave-one-out testing and internal n-fold cross-validation
Multi-threaded server functionality
Loading .....
Recently Viewed Categories
Recently Viewed Products
