v0.3.0

Gogeta v0.3.0

Diff since v0.2.0

Added partition-based neural network formulation and experimental support for input convex neural networks. Support for importing model parameters as JSON files. Improved use of NN surrogates with anonymous variables.
Improvements to the documentation and code examples.

Merged pull requests:

• Marina (#27) (@kankeinai)
• removed print statement from sampling.jl (#28) (@mamorim00)
• Icnns (#29) (@EetuReijonen)
• ICNNs and partition-split formulation (#31) (@EetuReijonen)

v0.2.0

Gogeta v0.2.0

Diff since v0.1.0

Added relaxing walk heuristic optimization algorithm and improved documentation.

Merged pull requests:

• Relaxing walk and new version (#26) (@EetuReijonen)

v0.1.0

Gogeta v0.1.0

This is the first public version of the Gogeta.jl package.
It's still under extensive development meaning that new features will be added and some existing features may change.
The documentation is limited at the time but will be expanded in the future.

Please report issues and development ideas to the Gogeta.jl GitHub page.

Merged pull requests:

• Tree ensemble code refactorization (#15) (@EetuReijonen)
• Rewrite neural network code (#18) (@EetuReijonen)
• Package finalization (#19) (@EetuReijonen)
• Convolutional neural network (CNN) functionality (#20) (@EetuReijonen)
• Solver independence refactor (#21) (@EetuReijonen)
• Sampling and package improvements (#22) (@EetuReijonen)

Closed issues:

• Registration (#23)