NEWS

---

Modeler 3.4.3 (2018-05-18)

UPDATES

• Changed use of S4 classes to avoid deprecated or discourged elements.

Modeler 3.4.2 (2017-07-13)

DOCUMENTATION

• Filled in most of the gaps inthe man pages.

Modeler 3.4.0

NEW FEATURES

• Added a separate neural net model.

Modeler 3.3.14

ADMINISTRATIVE

• FIxed NAMESPACE issues.

Modeler 2.3.0

ADMINISTRATIVE

• Changed LICENSE to Apache 2.0

Modeler 2.2.1

BUG FIXES

• Made compatible with new R check steps that verify DESCRIPTION and NAMESPACE dependencies.

Modeler 2.2.0

NEW FEATURES

• Added objects and functions for the following methods: support vector machines (SVM), neual networks (NNET), and random forests (RF).

Modeler 2.1.0

DEPRECATED

• Instead of exporting the paired functions, we replaced each set with a precomputed object of the Modeler class that already combines each pair. The old functions still exist, but are planned to be removed from the namespace in a future version. If you need to call them directly, then you should fully qualify them with the class name (e.g., Modeler:::learnLR).

Modeler 2.0.0

NEW FEATURES

• Brand new packaging: split out from the CrossVal package. Contains the generic support to "learn" and "predict". Also contains the base implementations of numerous statistical models for classification or regression.