About: Crino: a neural-network library based on Theano

1.0.0 (7 july 2014) : - Initial release of crino - Implements a torch-like library to build artificial neural networks (ANN) - Provides standard implementations for : * auto-encoders * multi-layer perceptrons (MLP) * deep neural networks (DNN) * input output deep architecture (IODA) - Provides a batch-gradient backpropagation algorithm, with adaptative learning rate

About: ARTOS can be used to quickly learn models for visual object detection without having to collect a set of samples manually. To make this possible, it uses ImageNet, a large image database with more than 20,000 categories.

Initial Announcement on mloss.org.

About: PyStruct is a framework for learning structured prediction in Python. It has a modular interface, similar to the well-known SVMstruct. Apart from learning algorithms it also contains model formulations for popular CRFs and interfaces to many inference algorithm implementation.

Initial Announcement on mloss.org.

About: Universal Python-written numerical optimization toolbox. Problems: NLP, LP, QP, NSP, MILP, LSP, LLSP, MMP, GLP, SLE, MOP etc; general logical constraints, categorical variables, automatic differentiation, stochastic programming, interval analysis, many other goodies

http://openopt.org/Changelog

About: peewit provides services for programming, running and result examination of machine learning experiments. It does not include any ML algorithms, has no GUI, and presumes certain uniformity of the experimental layout. But it does not make assumptions on the type of task under study. The current version-number is 0.10.

v-cube with side-cubes

About: Jubatus is a general framework library for online and distributed machine learning. It currently supports classification, regression, clustering, recommendation, nearest neighbors, anomaly detection, and graph analysis. Loose model sharing provides higher scalability, better performance, and real-time capabilities, by combining online learning with distributed computations.

0.5.0 add new supports for clustering and nearest neighbors. For more detail, see http://t.co/flMcTcYZVs

About: Bob is a free signal-processing and machine learning toolbox originally developed by the Biometrics group at Idiap Research Institute, in Switzerland.

Bob 1.2.0 comes about 1 year after we released Bob 1.0.0. This new release comes with a big set of new features and lots of changes under the hood to make your experiments run even smoother. Some statistics:

Diff URL: https://github.com/idiap/bob/compare/v1.1.4...HEAD Commits: 629 Files changed: 954 Contributors: 7

Here is a quick list of things you should pay attention for while integrating your satellite packages against Bob 1.2.x:

• The LBP module had its API changed look at the online docs for more details
• LLRTrainer has been renamed to CGLogRegTrainer
• The order in which you pass data to CGLogRegTrainer has been inverted (negatives now go first)
• For C++ bindings, includes are in bob/python instead of bob/core/python
• All specialized Bob exceptions are gone, if you were catching them, most have been cast into std::runtime_error's

For a detailed list of changes and additions, please look at our Changelog page for this release and minor updates:

https://github.com/idiap/bob/wiki/Changelog-from-1.1.4-to-1.2 https://github.com/idiap/bob/wiki/Changelog-from-1.2.0-to-1.2.1 https://github.com/idiap/bob/wiki/Changelog-from-1.2.1-to-1.2.2

About: Embarrassingly Parallel Array Computing: EPAC is a machine learning workflow builder.

Initial Announcement on mloss.org.

About: MyMediaLite is a lightweight, multi-purpose library of recommender system algorithms.

Mostly bug fixes.

For details see: https://github.com/zenogantner/MyMediaLite/blob/master/doc/Changes

About: ClowdFlows is a web based platform for service oriented data mining publicly available at http://clowdflows.org . A web based interface allows users to construct data mining workflows that are hosted on the web and can be (if allowed by the author) accessed by anyone by following a URL of the workflow.

Initial Announcement on mloss.org.

About: A library for artificial neural networks.

Added algorithms:

• L-BFGS optimizer
• k-means
• sparse auto-encoder
• preprocessing: normalization, PCA, ZCA whitening

About: This package includes implementations of the CCM, DMV and DMV+CCM parsers from Klein and Manning (2004), and code for testing them with the WSJ, Negra and Cast3LB corpuses (English, German and Spanish respectively). A detailed description of the parsers can be found in Klein (2005).

Initial Announcement on mloss.org.

About: Test submission. Is MLOSS working?

Initial Announcement on mloss.org.

About: The mission of this project is to build and support a community interested in machine learning and machine intelligence based on modeling the neocortex and the principles upon which it works.

Initial Announcement on mloss.org.

About: A collection of python code to perform research in optimization. The aim is to provide reusable components that can be quickly applied to machine learning problems. Used in: - Ellipsoidal multiple instance learning - difference of convex functions algorithms for sparse classfication - Contextual bandits upper confidence bound algorithm (using GP) - learning output kernels, that is kernels between the labels of a classifier.

• minor bugfix

About: A descriptive and programming language independent format and API for the simplified configuration, documentation, and design of computer experiments.

Initial Announcement on mloss.org.

About: HDDM is a python toolbox for hierarchical Bayesian parameter estimation of the Drift Diffusion Model (via PyMC). Drift Diffusion Models are used widely in psychology and cognitive neuroscience to study decision making.

• New and improved HDDM model with the following changes:
  • Priors: by default model will use informative priors (see http://ski.clps.brown.edu/hddm_docs/methods.html#hierarchical-drift-diffusion-models-used-in-hddm) If you want uninformative priors, set informative=False.
  • Sampling: This model uses slice sampling which leads to faster convergence while being slower to generate an individual sample. In our experiments, burnin of 20 is often good enough.
  • Inter-trial variablity parameters are only estimated at the group level, not for individual subjects.
  • The old model has been renamed to HDDMTransformed.
  • HDDMRegression and HDDMStimCoding are also using this model.
• HDDMRegression takes patsy model specification strings. See http://ski.clps.brown.edu/hddm_docs/howto.html#estimate-a-regression-model and http://ski.clps.brown.edu/hddm_docs/tutorial_regression_stimcoding.html#chap-tutorial-hddm-regression
• Improved online documentation at http://ski.clps.brown.edu/hddm_docs
• A new HDDM demo at http://ski.clps.brown.edu/hddm_docs/demo.html
• Ratcliff's quantile optimization method for single subjects and groups using the .optimize() method
• Maximum likelihood optimization.
• Many bugfixes and better test coverage.
• hddm_fit.py command line utility is depracated.

About: Orange is a component-based machine learning and data mining software. It includes a friendly yet powerful and flexible graphical user interface for visual programming. For more advanced use(r)s, [...]

The core of the system (except the GUI) no longer includes any GPL code and can be licensed under the terms of BSD upon request. The graphical part remains under GPL.

Changed the BibTeX reference to the paper recently published in JMLR MLOSS.

About: Python Framework for Vector Space Modelling that can handle unlimited datasets (streamed input, online algorithms work incrementally in constant memory).

• added the "hashing trick" (by Homer Strong)
• support for adding target classes in SVMlight format (by Corrado Monti)
• fixed problems with global lemmatizer object when running in parallel on Windows
• parallelization of Wikipedia processing + added script version that lemmatizes the input documents
• added class method to initialize Dictionary from an existing corpus (by Marko Burjek)

About: Python Machine Learning Toolkit

Added LASSO (using coordinate descent optimization). Made SVM classification (learning and applying) much faster: 2.5x speedup on yeast UCI dataset.