Encog is an advanced machine learning framework that supports a variety of advanced algorithms, as well as support classes to normalize and process data, its training algoritms are multi-threaded and scale well to multicore hardware and can also make use of a GPU to further speed processing time. A GUI based workbench is also provided to help model and train machine learning algorithms.
Customer Reviews
G2 User in Research
Advanced user of Encog Machine Learning FrameworkIt is open-source software and does not require license or activation fees. I like that it came with support for several different Neural Network topologies and training techniques. As far as Java-based computational frameworks are concerned, the performance of the library was acceptable.
I found the learning curve for customisations to be quite steep; in particular, plugging the framework into a custom visualisation platform required a significant amount of development. If you're looking for more novel fitness definitions, such as altruistic performance in a co-operative task, you may need to make extensive changes within the library itself, instead of simply supplying the library with the necessary parameters.
Consider TensorFlow instead, unless you absolutely must develop in Java.
At time of use, Encog was a platform written in Java, a programming language that all the members of my team were familiar with. This heavily skewed the selection process. We discovered that heavy alterations to the algorithms we intended to use was required for our research objectives. In the end, the ability to build upon the open-source library to realise our research goals meant that we were able to carry out our Evolutionary Computing research successfully.
