The nnlib2 library and nnlib2Rcpp R package for implementing neural networks - v.0.2.9

Changes to nnlib2Rcpp version 0.2.9 (from 0.2.8)

minor changes in documentation of supervised LVQ (LVQs) helper functions.
support importing a set of codebook vectors in LVQs_train function (to continue training from previous LVQs_train invocations, use custom initial weights etc).

Scientific Software - Peer-reviewed - C++
Published by VNNikolaidis about 1 year ago

The nnlib2 library and nnlib2Rcpp R package for implementing neural networks - v.0.2.8

Changes to nnlib2Rcpp version 0.2.8 (from 0.2.7)

minor change in documentation of LVQ, explaining why data should be in 0 to 1 range if random values in [0 1] are used for initial weights.
added (due to recent supervised LVQ popularity) two LVQs wrapper functions to simplify using LVQs (functions LVQs_train, LVQs_recall).

Scientific Software - Peer-reviewed - C++
Published by VNNikolaidis over 1 year ago

The nnlib2 library and nnlib2Rcpp R package for implementing neural networks - v.0.2.7

Minor change to eliminate a minor issue reported on CRAN additional checks. Does not affect functionality.

Scientific Software - Peer-reviewed - C++
Published by VNNikolaidis almost 2 years ago

The nnlib2 library and nnlib2Rcpp R package for implementing neural networks - v.0.2.6

Changes to nnlib2Rcpp version 0.2.6 (from 0.2.5) - changes to LVQs (and related lvqnn, lvqconnectionset classes) to make it more fit for use on real data: allow lvqnn getting and setting limits for weight values, define training coefficients etc. - added related methods to LVQs (setweightlimits, setencodingcoefficients). - removed directive (usepackage[english]{babel}) from vignette file ('intro.ltx'). - added trainsingle method (encodes a single [data vector,class] pair). Used in combination with setencoding_coefficients, allows allowing different training coefficients (reward, punish) per class (this may be useful in highly unbalanced class problems). - added option to LVQs recall (classify) data ignoring nodes that had less than a number of rewards received during encoding. - changed LVQs examples. - other minor changes.

Scientific Software - Peer-reviewed - C++
Published by VNNikolaidis almost 2 years ago

The nnlib2 library and nnlib2Rcpp R package for implementing neural networks - v.0.2.5

Changes to nnlib2Rcpp version 0.2.5 (from 0.2.4)

added example in NN documentation (NN-class.Rd).
changes resulting from discussion in https://github.com/VNNikolaidis/nnlib2Rcpp/issues/16:
added get_weights, set_weights methods to LVQs (allows access to codebook vectors).
added set_number_of_nodes_per_class method to LVQs (allows multiple output nodes (i.e codebook vectors) per class in the training data).
added get_number_of_nodes_per_class method to LVQs (returns number of output nodes that are used per class).
added get_number_of_rewards method to LVQs (counts number of times each output node was positively reinforced).
added setup method to LVQs (allows setting up an untrained supervised LVQ).
added enable_punishment, disable_punishment methods to LVQs (enable/disable negative reinforcement).
other minor changes.

Scientific Software - Peer-reviewed - C++
Published by VNNikolaidis almost 2 years ago

The nnlib2 library and nnlib2Rcpp R package for implementing neural networks - v.0.2.4

Changes to nnlib2Rcpp version 0.2.3 (from 0.2.2)

minor change in documentation (nnlib2Rcpp.Rd, as suggested by Kurt Hornik)

Changes to nnlib2Rcpp version 0.2.4 (from 0.2.3)

minor corrections in documentation (NN-R-components.Rd)
in documentation, file examples which employ the matrix multiplication operator (%*%), the percent symbol is replaced by backslash-percent so that these examples appear correctly in the help viewer.
in documentation, changed all files to eliminate "Unnecessary braces" notes when tools::checkRd() is used.

Scientific Software - Peer-reviewed - C++
Published by VNNikolaidis about 2 years ago

The nnlib2 library and nnlib2Rcpp R package for implementing neural networks - v.0.2.2

Changes to nnlib2Rcpp version 0.2.2 (from 0.2.1)

minor corrections and changes in documentation.
added 'display_rate' parameter to Autoencoder function (controls or limits display of current error when training).
added 'mute' method to 'BP' module (enables/disables display of current error when training).

Scientific Software - Peer-reviewed - C++
Published by VNNikolaidis almost 3 years ago

The nnlib2 library and nnlib2Rcpp R package for implementing neural networks - v.0.2.1

Changes to nnlib2Rcpp version 0.2.1 (from 0.2.0)

(nnlib2) added class 'generic_connection_matrix', a matrix-based connection set that can be used to replace generic_connection_set (this is derived from connection_set, working draft, needs improvements, expect small run time performance gains).
(nnlib2) used generic_connection_matrix derived classes in BP and related neural nets. Note: Option #define BP_CONNECTIONS in 'nn_bp.cpp' selects whether bp_connection_set (based on 'generic_connection_set') or bp_connection_matrix (based on generic_connection_matrix) is to be used.
added example_connection_matrix (example that creates a matrix-based connection set to be used in NN module via its add_connection_set).
Rcpp_aux_control_R.h was renamed to Rcpp_R_aux_control.h.
modified "add_connection_set" method of module "NN" to also accept a list (R object) containing the parameters. This also allows for user-defined parameters to be passed when creating new connection sets.
modified "connect_layers_at" and "fully_connect_layers_at" methods of module "NN" to also accept a list (R object) containing the parameters for creating the new connection set. This also allows for user-defined parameters to be passed when creating new layers.
modified "add_layer" method of module "NN" to also accept a list (R object) containing the parameters. This also allows for user-defined parameters to be passed when creating new layers.
added 'R-connections' (C++ class 'R_connection_matrix'), a matrix-based connection_set type that calls R functions when encoding/recalling (file Rcpp_R_layer.h).
added 'R-layer' (C++ class 'R_layer') a layer type that calls R functions when encoding/recalling (file Rcpp_R_connection_matrix.h).
added support and documentation (file NN-R-components.Rd) about using 'R-layer' and 'R-connections' in 'NN' module.
other minor changes.

Scientific Software - Peer-reviewed - C++
Published by VNNikolaidis almost 3 years ago

The nnlib2 library and nnlib2Rcpp R package for implementing neural networks - v.0.2.0.a

Fixed minor issues to release to CRAN

Scientific Software - Peer-reviewed - C++
Published by VNNikolaidis about 3 years ago

The nnlib2 library and nnlib2Rcpp R package for implementing neural networks - v.0.2.0

Changes to nnlib2Rcpp version 0.2.0 (from 0.1.10) - modified handling of errors and warnings. - added method to module "NN" (getmiscvaluesat) - added method to module "NN" (gettopologyinfo, returns data.frame with NN topology information). - added "softmax", "bp-hidden-softmax" and "bp-output-softmax" layer types (see https://github.com/VNNikolaidis/nnlib2Rcpp/issues/14) - added documentation describing the names of available NN components (NNcomponentnames.Rd). - added method to module "NN" (setinputat (same as inputat)). - split additionalparts.h to several files (for less clutter when users create their components). - small run time performance improvements (less checks in dllist operations). - (bug fix) added call to inputfunction() in pe moveinputtooutput(). This allows passthroughlayer to receive input from connections (and not just directly via setinput). - enabled NN's getinputat() method for layers (there are issues in the way this is currently implemented, see code for "previewcurrentinput" method of pe c++ class). - added "setweightsat" method to module "NN". - added auxcontrol component for calling R functions from the NN (RcppauxcontrolR.h). - added methods to module "NN" ("addRforwarding", "addRpipelining", "addRignoring" and "addRfunction"). These methods create NN components that can call an R function during NN processing (encoding or recalling data). - added options to Autoencoder function that allow to stop training when an acceptable error level has been reached. - added option to module "BP" (method "seterrorlevel") to stop training when an acceptable error level has been reached. - added to module "NN" methods "encodeallfwd", "encodeallbwd", "recallallfwd" and "recallallbwd". - other minor changes.

Scientific Software - Peer-reviewed - C++
Published by VNNikolaidis about 3 years ago

The nnlib2 library and nnlib2Rcpp R package for implementing neural networks - v.0.1.11

Changes to nnlib2Rcpp version 0.1.11 (from 0.1.10)

modified handling of errors and warnings.
added method to module "NN" (get_misc_values_at)
added method to module "NN" (get_topology_info, returns data.frame with NN topology information).
added "softmax", "bp-hidden-softmax" and "bp-output-softmax" layer types.
added documentation describing the names of available NN components (NN_component_names.Rd)
(Note: v.0.1.11 is not released to CRAN)

Scientific Software - Peer-reviewed - C++
Published by VNNikolaidis about 3 years ago

The nnlib2 library and nnlib2Rcpp R package for implementing neural networks - v.0.1.10

Changes to nnlib2Rcpp version 0.1.10 (from 0.1.9) - NN method addlayer now accepts additional optional parameter (double) to be used by specific layer implementations (for example, BP layers interpret it is as the learning rate). More optional parameters may be added in future versions. - NN methods addconnectionset, connectlayersat and fullyconnectlayersat now accept additional optional parameter (double) to be used by specific connection set implementations (for example, BP connection sets interpret it is as the learning rate). More optional parameters may be added in future versions. - includes nnlib2 v.0.2.6 (added functionality to directly get/set biases in layer processing elements (PEs)). - changes in NN module to support get/set biases (added functions: setbiasesat, setbiasat, getbiasesat, getbiasat). - numbers specifying PE and connections in NN module functions now start from 0 (was 1, changed to agree with those shown when NN is printed. This affects functions: getweightat, setweightat and removesingleconnection). - other minor changes.

Scientific Software - Peer-reviewed - C++
Published by VNNikolaidis about 3 years ago

The nnlib2 library and nnlib2Rcpp R package for implementing neural networks - v.0.1.9

Minor changes

Scientific Software - Peer-reviewed - C++
Published by VNNikolaidis almost 4 years ago

The nnlib2 library and nnlib2Rcpp R package for implementing neural networks - v.0.1.8

Minor documentation changes. Other changes include:

What's Changed

• Define STRICTRHEADERS, include float.h for DBL_MIN by @eddelbuettel in https://github.com/VNNikolaidis/nnlib2Rcpp/pull/11
• Minor formatting changes. by @arfon in https://github.com/VNNikolaidis/nnlib2Rcpp/pull/12

New Contributors

• @eddelbuettel made their first contribution in https://github.com/VNNikolaidis/nnlib2Rcpp/pull/11
• @arfon made their first contribution in https://github.com/VNNikolaidis/nnlib2Rcpp/pull/12

Full Changelog: https://github.com/VNNikolaidis/nnlib2Rcpp/compare/v0.1.5...v.0.1.8

Scientific Software - Peer-reviewed - C++
Published by VNNikolaidis over 4 years ago

The nnlib2 library and nnlib2Rcpp R package for implementing neural networks - v.0.1.7

Minor changes (only affect source) due to support Rcpp changes. Released to be consistent with version on CRAN) Re-released for access from Zenodo open-access repository (no code changes).

Scientific Software - Peer-reviewed - C++
Published by VNNikolaidis almost 5 years ago

The nnlib2 library and nnlib2Rcpp R package for implementing neural networks - v0.1.6

Changes to nnlib2Rcpp version 0.1.6 (from 0.1.5)

added method to module "NN" (recall_dataset) to decode/map/retrieve output for entire input dataset.
added method to module "NN" (encode_dataset_unsupervised) for faster unsupervised training.
added method to module "NN" (encode_datasets_supervised) for faster supervised training.
includes nnlib2 v.0.2.4
added method to module "NN" (get_output_at).
added method to module "NN" (set_output_at).
added package vingette (also found in doc/manual.pdf)
other minor changes

Scientific Software - Peer-reviewed - C++
Published by VNNikolaidis almost 5 years ago

The nnlib2 library and nnlib2Rcpp R package for implementing neural networks - v0.1.5

Changes to nnlib2Rcpp version 0.1.5 (from 0.1.4) - added some methods to module "NN" (getweightat, setweightat) - added some methods to module "NN" (setmiscvalues_at)

Scientific Software - Peer-reviewed - C++
Published by VNNikolaidis over 5 years ago