All entries.
Showing Items 21-30 of 658 on page 3 of 66: Previous 1 2 3 4 5 6 7 8 Next Last

Logo Spectra. A Library for Large Scale Eigenvalue Problems 0.5.0

by yixuanq - September 13, 2017, 02:34:21 CET [ Project Homepage BibTeX Download ] 772 views, 245 downloads, 2 subscriptions

About: A header-only C++ library for solving large scale eigenvalue problems

Changes:

Initial Announcement on mloss.org.


About: Tool aimed at helping remedy the reproducibility problem, specifically in the statistical and data wrangling aspects.

Changes:

Initial Announcement on mloss.org.


Logo Top Frequency Based Parallel Coordinates 1.0.0

by matloff - September 5, 2017, 05:49:41 CET [ Project Homepage BibTeX BibTeX for corresponding Paper Download ] 871 views, 172 downloads, 1 subscription

About: A novel method to create parallel coordinates plots on large data sets without causing a "black screen" problem.

Changes:

Initial Announcement on mloss.org.


Logo JMLR MLPACK 2.2.5

by rcurtin - August 26, 2017, 06:07:47 CET [ Project Homepage BibTeX BibTeX for corresponding Paper Download ] 94268 views, 16870 downloads, 6 subscriptions

Rating Whole StarWhole StarWhole StarWhole Star1/2 Star
(based on 1 vote)

About: A scalable, fast C++ machine learning library, with emphasis on usability.

Changes:

Released August 25, 2017.

  • Compilation fix for some systems (#1082).

  • Fix PARAM_INT_OUT() (#1100).


Logo python weka wrapper3 0.1.3

by fracpete - August 23, 2017, 01:18:36 CET [ Project Homepage BibTeX Download ] 4457 views, 978 downloads, 3 subscriptions

About: A thin Python3 wrapper that uses the javabridge Python library to communicate with a Java Virtual Machine executing Weka API calls.

Changes:
  • added check_for_modified_class_attribute method to FilterClassifier class
  • added complete_classname method to weka.core.classes module, which allows completion of partial classnames like .J48 to weka.classifiers.trees.J48; if there is a unique match; JavaObject.new_instance and JavaObject.check_type now make use of this functionality, allowing for instantiations like Classifier(cls=".J48")
  • jvm.start(system_cp=True) no longer fails with a KeyError: 'CLASSPATH' if there is no CLASSPATH environment variable defined
  • Libraries mtl.jar, core.jar and arpack_combined_all.jar were added as is to the weka.jar in the 3.9.1 release instead of adding their content to it. Repackaged weka.jar to fix this issue.

Logo python weka wrapper 0.3.11

by fracpete - August 23, 2017, 01:17:24 CET [ Project Homepage BibTeX Download ] 55119 views, 11048 downloads, 3 subscriptions

About: A thin Python wrapper that uses the javabridge Python library to communicate with a Java Virtual Machine executing Weka API calls.

Changes:
  • added check_for_modified_class_attribute method to FilterClassifier class
  • added complete_classname method to weka.core.classes module, which allows completion of partial classnames like .J48 to weka.classifiers.trees.J48; if there is a unique match; JavaObject.new_instance and JavaObject.check_type now make use of this functionality, allowing for instantiations like Classifier(cls=".J48")
  • jvm.start(system_cp=True) no longer fails with a KeyError: 'CLASSPATH' if there is no CLASSPATH environment variable defined
  • Libraries mtl.jar, core.jar and arpack_combined_all.jar were added as is to the weka.jar in the 3.9.1 release instead of adding their content to it. Repackaged weka.jar to fix this issue.

About: A non-iterative, incremental and hyperparameter-free learning method for one-layer feedforward neural networks without hidden layers. This method efficiently obtains the optimal parameters of the network, regardless of whether the data contains a greater number of samples than variables or vice versa. It does this by using a square loss function that measures errors before the output activation functions and scales them by the slope of these functions at each data point. The outcome is a system of linear equations that obtain the network's weights and that is further transformed using Singular Value Decomposition.

Changes:

Initial Announcement on mloss.org.


Logo r-cran-CORElearn 1.51.2

by r-cran-robot - August 8, 2017, 00:00:00 CET [ Project Homepage BibTeX Download ] 22053 views, 4396 downloads, 2 subscriptions

About: Classification, Regression and Feature Evaluation

Changes:

Fetched by r-cran-robot on 2017-12-01 00:00:05.182003


Logo KeLP 2.2.1

by kelpadmin - August 7, 2017, 17:20:39 CET [ Project Homepage BibTeX BibTeX for corresponding Paper Download ] 19979 views, 4194 downloads, 3 subscriptions

About: Kernel-based Learning Platform (KeLP) is Java framework that supports the implementation of kernel-based learning algorithms, as well as an agile definition of kernel functions over generic data representation, e.g. vectorial data or discrete structures. The framework has been designed to decouple kernel functions and learning algorithms, through the definition of specific interfaces. Once a new kernel function has been implemented, it can be automatically adopted in all the available kernel-machine algorithms. KeLP includes different Online and Batch Learning algorithms for Classification, Regression and Clustering, as well as several Kernel functions, ranging from vector-based to structural kernels. It allows to build complex kernel machine based systems, leveraging on JSON/XML interfaces to instantiate prediction models without writing a single line of code.

Changes:

In addition to minor bug fixes, this release includes:

  • A new cache (FixSizeKernelCache) that can store a larger number of computations.

  • Evaluators for measuring the quality of Clustering algorithms.

Furthermore we also released the new module kelp-input-generator, that contains the facilities to parse text snippets and generate tree representations for KeLP!

Check out this new version from our repositories. API Javadoc is already available. Your suggestions will be very precious for us, so download and try KeLP 2.2.1!


About: A non-iterative learning method for one-layer (no hidden layer) neural networks, where the weights can be calculated in a closed-form manner, thereby avoiding low convergence rate and also hyperparameter tuning. The proposed learning method, LANN-SVD in short, presents a good computational efficiency for large-scale data analytic.

Changes:

Initial Announcement on mloss.org.


Showing Items 21-30 of 658 on page 3 of 66: Previous 1 2 3 4 5 6 7 8 Next Last