THIS SOURCE CODE IS SUPPLIED "AS IS" WITHOUT WAR- RANTY OF ANY KIND,
AND ITS AUTHOR AND THE JOURNAL OF MACHINE LEARNING RESEARCH (JMLR) AND
JMLR'S PUBLISH- ERS AND DISTRIBUTORS, DISCLAIM ANY AND ALL WARRANTIES,
INCLUDING BUT NOT LIMITED TO ANY IMPLIED WARRANTIES OF MERCHANTABILITY
AND FITNESS FOR A PARTICULAR PUR- POSE, AND ANY WARRANTIES OR NON
INFRINGEMENT. THE USER ASSUMES ALL LIABILITY AND RESPONSIBILITY FOR
USE OF THIS SOURCE CODE, AND NEITHER THE AUTHOR NOR JMLR, NOR JMLR'S
PUBLISHERS AND DISTRIBUTORS, WILL BE LIABLE FOR DAMAGES OF ANY KIND
RESULTING FROM ITS USE. Without limiting the generality of the
foregoing, neither the author, nor JMLR, nor JMLR's publishers and
distributors, warrant that the Source Code will be error-free, will
operate without interruption, or will meet the needs of the user.
-----------------------------------
The interesting program files are:

- TC4L2L.m, performing Task Clustering for Learning to Learn (multitask learning) </LI>
- TG4L2L.m, performing Task Gating for Learning to Learn </LI>
  
Examples of how the clustering and gating software is used, can be found in

TCSimulations.m and TGSimulations.m,

implementation and optimization of the task-dependent prior mean can be found in

TFSimulations


After optimization, predictions for future outputs and (for testing) explained
variance on future data can be obtained through:

- Pred4L2L.m  (Predictions for Learning to Learn through clustering)
- GPred4L2L.m (Predictions for Learning to Learn through Gating)
- FPred4L2L.m (Predictions for Learning to Learn through a task-dependent prior mean (using Features))

