Skip to content

Commit 96c0b25

Browse files
ahmedfgadahmedfgad
authored
PyGAD 2.7.0 Docs
1 parent 0342b13 commit 96c0b25

File tree

5 files changed

+629
-53
lines changed

5 files changed

+629
-53
lines changed

‎docs/source/Footer.rst

+76-17
Original file line numberDiff line numberDiff line change
@@ -320,6 +320,65 @@ Release Date: 6 August 2020
320320

321321
.. _header-n124:
322322

323+
PyGAD 2.7.0
324+
-----------
325+
326+
Changes in PyGAD 2.7.0 (11 September 2020):
327+
328+
1. The ``learning_rate`` parameter in the ``pygad.nn.train()`` function
329+
defaults to **0.01**.
330+
331+
2. Added support of building neural networks for regression using the
332+
new parameter named ``problem_type``. It is added as a parameter to
333+
both ``pygad.nn.train()`` and ``pygad.nn.predict()`` functions. The
334+
value of this parameter can be either **classification** or
335+
**regression** to define the problem type. It defaults to
336+
**classification**.
337+
338+
3. The activation function for a layer can be set to the string
339+
``"None"`` to refer that there is no activation function at this
340+
layer. As a result, the supported values for the activation function
341+
are ``"sigmoid"``, ``"relu"``, ``"softmax"``, and ``"None"``.
342+
343+
To build a regression network using the ``pygad.nn`` module, just do the
344+
following:
345+
346+
1. Set the ``problem_type`` parameter in the ``pygad.nn.train()`` and
347+
``pygad.nn.predict()`` functions to the string ``"regression"``.
348+
349+
2. Set the activation function for the output layer to the string
350+
``"None"``. This sets no limits on the range of the outputs as it
351+
will be from ``-infinity`` to ``+infinity``. If you are sure that all
352+
outputs will be nonnegative values, then use the ReLU function.
353+
354+
Check the documentation of the ``pygad.nn`` module for an example that
355+
builds a neural network for regression. The regression example is also
356+
available at `this GitHub
357+
project <https://github.com/ahmedfgad/NumPyANN>`__:
358+
https://github.com/ahmedfgad/NumPyANN
359+
360+
To build and train a regression network using the ``pygad.gann`` module,
361+
do the following:
362+
363+
1. Set the ``problem_type`` parameter in the ``pygad.nn.train()`` and
364+
``pygad.nn.predict()`` functions to the string ``"regression"``.
365+
366+
2. Set the ``output_activation`` parameter in the constructor of the
367+
``pygad.gann.GANN`` class to ``"None"``.
368+
369+
Check the documentation of the ``pygad.gann`` module for an example that
370+
builds and trains a neural network for regression. The regression
371+
example is also available at `this GitHub
372+
project <https://github.com/ahmedfgad/NeuralGenetic>`__:
373+
https://github.com/ahmedfgad/NeuralGenetic
374+
375+
To build a classification network, either ignore the ``problem_type``
376+
parameter or set it to ``"classification"`` (default value). In this
377+
case, the activation function of the last layer can be set to any type
378+
(e.g. softmax).
379+
380+
.. _header-n148:
381+
323382
PyGAD Projects at GitHub
324383
========================
325384

@@ -328,7 +387,7 @@ https://pypi.org/project/pygad. PyGAD is built out of a number of
328387
open-source GitHub projects. A brief note about these projects is given
329388
in the next subsections.
330389

331-
.. _header-n126:
390+
.. _header-n150:
332391

333392
`GeneticAlgorithmPython <https://github.com/ahmedfgad/GeneticAlgorithmPython>`__
334393
--------------------------------------------------------------------------------
@@ -339,7 +398,7 @@ GitHub Link: https://github.com/ahmedfgad/GeneticAlgorithmPython
339398
is the first project which is an open-source Python 3 project for
340399
implementing the genetic algorithm based on NumPy.
341400

342-
.. _header-n129:
401+
.. _header-n153:
343402

344403
`NumPyANN <https://github.com/ahmedfgad/NumPyANN>`__
345404
----------------------------------------------------
@@ -353,7 +412,7 @@ neural network without using a training algorithm. Currently, it only
353412
supports classification and later regression will be also supported.
354413
Moreover, only one class is supported per sample.
355414

356-
.. _header-n132:
415+
.. _header-n156:
357416

358417
`NeuralGenetic <https://github.com/ahmedfgad/NeuralGenetic>`__
359418
--------------------------------------------------------------
@@ -366,7 +425,7 @@ projects
366425
`GeneticAlgorithmPython <https://github.com/ahmedfgad/GeneticAlgorithmPython>`__
367426
and `NumPyANN <https://github.com/ahmedfgad/NumPyANN>`__.
368427

369-
.. _header-n135:
428+
.. _header-n159:
370429

371430
`NumPyCNN <https://github.com/ahmedfgad/NumPyCNN>`__
372431
----------------------------------------------------
@@ -378,7 +437,7 @@ convolutional neural networks using NumPy. The purpose of this project
378437
is to only implement the **forward pass** of a convolutional neural
379438
network without using a training algorithm.
380439

381-
.. _header-n138:
440+
.. _header-n162:
382441

383442
`CNNGenetic <https://github.com/ahmedfgad/CNNGenetic>`__
384443
--------------------------------------------------------
@@ -390,7 +449,7 @@ convolutional neural networks using the genetic algorithm. It uses the
390449
`GeneticAlgorithmPython <https://github.com/ahmedfgad/GeneticAlgorithmPython>`__
391450
project for building the genetic algorithm.
392451

393-
.. _header-n141:
452+
.. _header-n165:
394453

395454
Submitting Issues
396455
=================
@@ -407,7 +466,7 @@ is not working properly or to ask for questions.
407466
If this is not a proper option for you, then check the **Contact Us**
408467
section for more contact details.
409468

410-
.. _header-n145:
469+
.. _header-n169:
411470

412471
Ask for Feature
413472
===============
@@ -424,7 +483,7 @@ to ahmed.f.gad@gmail.com.
424483

425484
Also check the **Contact Us** section for more contact details.
426485

427-
.. _header-n149:
486+
.. _header-n173:
428487

429488
Projects Built using PyGAD
430489
==========================
@@ -443,15 +502,15 @@ Within your message, please send the following details:
443502

444503
- Preferably, a link that directs the readers to your project
445504

446-
.. _header-n160:
505+
.. _header-n184:
447506

448507
For More Information
449508
====================
450509

451510
There are different resources that can be used to get started with the
452511
genetic algorithm and building it in Python.
453512

454-
.. _header-n162:
513+
.. _header-n186:
455514

456515
Tutorial: Implementing Genetic Algorithm in Python
457516
--------------------------------------------------
@@ -475,7 +534,7 @@ good resource to start with coding the genetic algorithm.
475534

476535
|image0|
477536

478-
.. _header-n173:
537+
.. _header-n197:
479538

480539
Tutorial: Introduction to Genetic Algorithm
481540
-------------------------------------------
@@ -494,7 +553,7 @@ which is available at these links:
494553

495554
|image1|
496555

497-
.. _header-n183:
556+
.. _header-n207:
498557

499558
Tutorial: Build Neural Networks in Python
500559
-----------------------------------------
@@ -514,7 +573,7 @@ available at these links:
514573

515574
|image2|
516575

517-
.. _header-n193:
576+
.. _header-n217:
518577

519578
Tutorial: Optimize Neural Networks with Genetic Algorithm
520579
---------------------------------------------------------
@@ -534,7 +593,7 @@ available at these links:
534593

535594
|image3|
536595

537-
.. _header-n203:
596+
.. _header-n227:
538597

539598
Tutorial: Building CNN in Python
540599
--------------------------------
@@ -560,7 +619,7 @@ good resource to start with coding CNNs.
560619

561620
|image4|
562621

563-
.. _header-n216:
622+
.. _header-n240:
564623

565624
Tutorial: Derivation of CNN from FCNN
566625
-------------------------------------
@@ -579,7 +638,7 @@ which is available at these links:
579638

580639
|image5|
581640

582-
.. _header-n226:
641+
.. _header-n250:
583642

584643
Book: Practical Computer Vision Applications Using Deep Learning with CNNs
585644
--------------------------------------------------------------------------
@@ -605,7 +664,7 @@ Find the book at these links:
605664
.. figure:: https://user-images.githubusercontent.com/16560492/78830077-ae7c2800-79e7-11ea-980b-53b6bd879eeb.jpg
606665
:alt:
607666

608-
.. _header-n241:
667+
.. _header-n265:
609668

610669
Contact Us
611670
==========

0 commit comments

Comments
 (0)