All entries.
Showing Items 121-130 of 535 on page 13 of 54: First Previous 8 9 10 11 12 13 14 15 16 17 18 Next Last

Logo ELKI 0.6.0

by erich - January 10, 2014, 18:32:28 CET [ Project Homepage BibTeX BibTeX for corresponding Paper Download ] 9073 views, 1688 downloads, 3 subscriptions

About: ELKI is a framework for implementing data-mining algorithms with support for index structures, that includes a wide variety of clustering and outlier detection methods.

Changes:

Additions and Improvements from ELKI 0.5.5:

Algorithms

Clustering:

  • Hierarchical Clustering - the slower naive variants were added, and the code was refactored
  • Partition extraction from hierarchical clusterings - different linkage strategies (e.g. Ward)
  • Canopy pre-Clustering
  • Naive Mean-Shift Clustering
  • Affinity propagation clustering (both with distances and similarities / kernel functions)
  • K-means variations: Best-of-multiple-runs, bisecting k-means
  • New k-means initialization: farthest points, sample initialization
  • Cheng and Church Biclustering
  • P3C Subspace Clustering
  • One-dimensional clustering algorithm based on kernel density estimation

Outlier detection

  • COP - correlation outlier probabilities
  • LDF - a kernel density based LOF variant
  • Simplified LOF - a simpler version of LOF (not using reachability distance)
  • Simple Kernel Density LOF - a simple LOF using kernel density (more consistent than LDF)
  • Simple outlier ensemble algorithm
  • PINN - projection indexed nearest neighbors, via projected indexes.
  • ODIN - kNN graph based outlier detection
  • DWOF - Dynamic-Window Outlier Factor (contributed by Omar Yousry)
  • ABOD refactored, into ABOD, FastABOD and LBABOD

Distances

  • Geodetic distances now support different world models (WGS84 etc.) and are subtantially faster.
  • Levenshtein distances for processing strings, e.g. for analyzing phonemes (contributed code, see "Word segmentation through cross-lingual word-to-phoneme alignment", SLT2013, Stahlberg et al.)
  • Bray-Curtis, Clark, Kulczynski1 and Lorentzian distances with R-tree indexing support
  • Histogram matching distances
  • Probabilistic divergence distances (Jeffrey, Jensen-Shannon, Chi2, Kullback-Leibler)
  • Kulczynski2 similarity
  • Kernel similarity code has been refactored, and additional kernel functions have been added

Database Layer and Data Types

Projection layer * Parser for simple textual data (for use with Levenshtein distance) Various random projection families (including Feature Bagging, Achlioptas, and p-stable) Latitude+Longitude to ECEF Sparse vector improvements and bug fixes New filter: remove NaN values and missing values New filter: add histogram-based jitter New filter: normalize using statistical distributions New filter: robust standardization using Median and MAD New filter: Linear discriminant analysis (LDA)

Index Layer

  • Another speed up in R-trees
  • Refactoring of M- and R-trees: Support for different strategies in M-tree New strategies for M-tree splits Speedups in M-tree
  • New index structure: in-memory k-d-tree
  • New index structure: in-memory Locality Sensitive Hashing (LSH)
  • New index structure: approximate projected indexes, such as PINN
  • Index support for geodetic data - (Details: Geodetic Distance Queries on R-Trees for Indexing Geographic Data, SSTD13)
  • Sampled k nearest neighbors: reference KDD13 "Subsampling for Efficient and Effective Unsupervised Outlier Detection Ensembles"
  • Cached (precomputed) k-nearest neighbors to share across multiple runs
  • Benchmarking "algorithms" for indexes

Mathematics and Statistics

  • Many new distributions have been added, now 28 different distributions are supported
  • Additional estimation methods (using advanced statistics such as L-Moments), now 44 estimators are available
  • Trimming and Winsorizing
  • Automatic best-fit distribution estimation
  • Preprocessor using these distributions for rescaling data sets
  • API changes related to the new distributions support
  • More kernel density functions
  • RANSAC covariance matrix builder (unfortunately rather slow)

Visualization

  • 3D projected coordinates (Details: Interactive Data Mining with 3D-Parallel-Coordinate-Trees, SIGMOD2013)
  • Convex hulls now also include nested hierarchical clusters

Other

  • Parser speedups
  • Sparse vector bug fixes and improvements
  • Various bug fixes
  • PCA, MDS and LDA filters
  • Text output was slightly improved (but still needs to be redesigned from scratch - please contribute!)
  • Refactoring of hierarchy classes
  • New heap classes and infrastructure enhancements
  • Classes can have aliases, e.g. "l2" for euclidean distance.
  • Some error messages were made more informative.
  • Benchmarking classes, also for approximate nearest neighbor search.

Logo JMLR Model Monitor 1.0

by traeder - August 17, 2009, 11:05:06 CET [ Project Homepage BibTeX BibTeX for corresponding Paper Download ] 12128 views, 1683 downloads, 0 comments, 1 subscription

About: Model Monitor is a Java toolkit for the systematic evaluation of classifiers under changes in distribution. It provides methods for detecting distribution shifts in data, comparing the performance [...]

Changes:

Improved AUROC calculation. Several minor bug fixes.


Logo SVQP 2

by leonbottou - January 31, 2009, 14:22:04 CET [ Project Homepage BibTeX Download ] 4992 views, 1674 downloads, 0 subscriptions

About: SVQP1 and SVQP2 are QP solvers for training SVM.

Changes:

Initial Announcement on mloss.org.


About: Infrastructure for representing, manipulating and analyzing transaction data and frequent patterns.

Changes:

Initial Announcement on mloss.org.


Logo ChemCpp 1.0.2

by pmahe - November 28, 2007, 21:47:44 CET [ Project Homepage BibTeX Download ] 4355 views, 1669 downloads, 0 subscriptions

Rating Whole StarWhole StarWhole Star1/2 StarEmpty Star
(based on 2 votes)

About: ChemCpp is a C++ toolbox for chemoinformatics focusing on the computation of kernel functions between chemical compounds.

Changes:

Initial Announcement on mloss.org.


Logo Piqle 2.0

by fdecomite - June 19, 2009, 10:16:53 CET [ Project Homepage BibTeX Download ] 3256 views, 1668 downloads, 1 subscription

About: Piqle (Platform for Implementing Q-Learning Experiments) is a Java framework for fast design, prototyping and test of reinforcement learning experiments (RL). By clearly separating algorithms and problems, it allows users to focus on either part of the RL paradigm:designing new algorithms or implementing new problems. Piqle implements many classical RL algorithms, making their parameters easily tunable. At this time, 13 problems are implemented, several with one or more variants. The user's manual explains in detail how to code a new problem. Written in Java, Piqle is as platform-independent as Java itself. Its components can easily be embedded as part of complex implementations, like robotics or decision making.

Changes:

Initial Announcement on mloss.org.


Logo BioSig for Octave and Matlab 2.31

by schloegl - July 28, 2009, 13:41:01 CET [ Project Homepage BibTeX BibTeX for corresponding Paper Download ] 10279 views, 1664 downloads, 0 subscriptions

About: BioSig is a software library for biomedical signal processings. Besides several other modules, one modul (t400) provides a common interface (train_sc.m and test_sc.m) to various classification [...]

Changes:

Update of project information: machine learning and classification tools are moved to the NaN-toolbox.


Logo JMLR PyBrain 0.3

by bayerj - March 3, 2010, 15:00:08 CET [ Project Homepage BibTeX BibTeX for corresponding Paper Download ] 14849 views, 1657 downloads, 2 subscriptions

About: PyBrain is a versatile machine learning library for Python. Its goal is to provide flexible, easy-to-use yet still powerful algorithms for machine learning tasks, including a variety of predefined [...]

Changes:
  • more documentation, including new tutorials
  • new and updated example scripts
  • major restructuring of the reinforcement learning part
  • homogeneous interface for optimization algorithms
  • fast networks (arac) are now in an independent package
  • new algorithms, network structures, tools...

Logo PREA Personalized Recommendation Algorithms Toolkit 1.1

by srcw - September 1, 2012, 22:53:37 CET [ Project Homepage BibTeX Download ] 6262 views, 1643 downloads, 2 subscriptions

About: An open source Java software providing collaborative filtering algorithms.

Changes:

Initial Announcement on mloss.org.


Logo MALLET 2.0-rc4

by jacktanner - August 24, 2009, 23:10:14 CET [ Project Homepage BibTeX BibTeX for corresponding Paper Download ] 10142 views, 1620 downloads, 1 subscription

About: MALLET is a Java-based package for statistical natural language processing, document classification, clustering, topic modeling, information extraction, and other machine learning applications to [...]

Changes:

MALLET 2.0 RC4 Release Notes July 16, 2009

Major updates:

An implementation of generalized expectation criteria training of MaxEnt classifiers and methods for obtaining constraints (c.f. Gregory Druck, Gideon Mann, Andrew McCallum "Learning from Labeled Features using Generalized Expectation Criteria.")

PagedInstanceList has been substantially rewritten by Mike Bond.

Bug fixes to topic model hyperparameter optimization and topic inference.


Showing Items 121-130 of 535 on page 13 of 54: First Previous 8 9 10 11 12 13 14 15 16 17 18 Next Last