If you are lucky enough to have one of the new Apple computers with ARM processors, you
might want to try our new WEKA downloads for this platform, which are available at the
(Of course, you can also run the Intel version of WEKA on ARM Macs by using Rosetta 2, but
that should be less efficient than code built specifically for this platform.)
Note that you may have to give the WEKA application permission to access relevant folders
on your Mac so that you can actually read data into WEKA. This seems to be a feature of
recent versions of macOS.
For fast native matrix algebra, we have also made a corresponding
package for WEKA. Installing this package using the WEKA package manager will give a speed
boost to those algorithms that use netlib-java under the hood. A list of some relevant
algorithms is appended below.
WEKA packages that DO NOT currently work with this distribution of WEKA: RPlugin (no
official distribution of R for Apple ARM yet), wekaDeeplearning4j (same), and Auto-WEKA
(the latter does not work with recent WEKA distributions anyway).
In contrast, it is possible to use wekaPython (for running scikit-learn schemes in WEKA)
by installing the scikit-learn package and other packages using miniforge3 and specifying
the path to the corresponding Python executable and directory in WEKA (or via environment
variables). The distributedWekaSpark3Dev and massiveOnlineAnalysis packages for WEKA also
seem to work fine.
Initial experience indicates that Apple ARM is a very good platform for running WEKA.
Execution times for algorithms are very good, and the user interface feels very snappy.
Here is the list of core WEKA schemes that directly or indirectly make use of
There are also some schemes in various packages: