Portability | portable |
---|---|
Stability | experimental |
Maintainer | amy@nualeargais.ie |
Safe Haskell | Safe-Inferred |
A modified Kohonen Self-organising Map (SOM) which supports a
time-independent learning function. (See
for a description of a SOM.)
SOM
References:
- Rougier, N. & Boniface, Y. (2011). Dynamic self-organising map. Neurocomputing, 74 (11), 1840-1847.
- Kohonen, T. (1982). Self-organized formation of topologically correct feature maps. Biological Cybernetics, 43 (1), 59–69.
- data DSOM gm k p
- defaultDSOM :: (Eq (Metric p), Ord (Metric p), Floating (Metric p)) => gm p -> Metric p -> Metric p -> DSOM gm k p
- customDSOM :: gm p -> (Metric p -> Metric p -> Metric p -> Metric p) -> DSOM gm k p
- rougierLearningFunction :: (Eq a, Ord a, Floating a) => a -> a -> a -> a -> a -> a
- toGridMap :: GridMap gm p => DSOM gm k p -> gm p
- trainNeighbourhood :: (Pattern p, FiniteGrid (gm p), GridMap gm p, Num (Metric p), Ord k, k ~ Index (gm p), k ~ Index (BaseGrid gm p), Fractional (Metric p)) => DSOM gm t p -> k -> p -> DSOM gm k p
Construction
A Self-Organising Map (DSOM).
Although DSOM
implements GridMap
, most users will only need the
interface provided by Data.Datamining.Clustering.Classifier
. If
you chose to use the GridMap
functions, please note:
- The functions
adjust
, andadjustWithKey
do not increment the counter. You can do so manually withincrementCounter
. - The functions
map
andmapWithKey
are not implemented (they just return anerror
). It would be problematic to implement them because the input DSOM and the output DSOM would have to have the sameMetric
type.
(GridMap gm p, ~ * k (Index (BaseGrid gm p)), Pattern p, FiniteGrid (gm p), GridMap gm (Metric p), ~ * k (Index (gm p)), ~ * k (Index (BaseGrid gm (Metric p))), Ord k, Ord (Metric p), Num (Metric p), Fractional (Metric p)) => Classifier (DSOM gm) k p | |
Foldable gm => Foldable (DSOM gm k) | |
(Foldable gm, GridMap gm p, FiniteGrid (BaseGrid gm p)) => GridMap (DSOM gm k) p | |
Grid (gm p) => Grid (DSOM gm k p) |
defaultDSOM :: (Eq (Metric p), Ord (Metric p), Floating (Metric p)) => gm p -> Metric p -> Metric p -> DSOM gm k pSource
Creates a classifier with a default (bell-shaped) learning
function. Usage is
, where:
defaultDSOM
gm r w t
gm
- The geometry and initial models for this classifier.
A reasonable choice here is
, wherelazyGridMap
g psg
is a
, andHexHexGrid
ps
is a set of random patterns. r
- and [
p
] are the first two parameters to the
.rougierLearningFunction
customDSOM :: gm p -> (Metric p -> Metric p -> Metric p -> Metric p) -> DSOM gm k pSource
Creates a classifier with a custom learning function.
Usage is
, where:
customDSOM
gm g
gm
- The geometry and initial models for this classifier.
A reasonable choice here is
, wherelazyGridMap
g psg
is a
, andHexHexGrid
ps
is a set of random patterns. f
- A function used to determine the learning rate (for adjusting the models in the classifier). This function will be invoked with three parameters. The first parameter will indicate how different the BMU is from the input pattern. The second parameter indicates how different the pattern of the node currently being trained is from the input pattern. The third parameter is the grid distance from the BMU to the node currently being trained, as a fraction of the maximum grid distance. The output is the learning rate for that node (the amount by which the node's model should be updated to match the target). The learning rate should be between zero and one.
rougierLearningFunction :: (Eq a, Ord a, Floating a) => a -> a -> a -> a -> a -> aSource
Configures a learning function that depends not on the time, but
on how good a model we already have for the target. If the
BMU is an exact match for the target, no learning occurs.
Usage is
, where rougierLearningFunction
r pr
is the
maximal learning rate (0 <= r <= 1), and p
is the elasticity.
NOTE: When using this learning function, ensure that
abs . difference
is always between 0 and 1, inclusive. Otherwise
you may get invalid learning rates.
Deconstruction
toGridMap :: GridMap gm p => DSOM gm k p -> gm pSource
Extracts the grid and current models from the DSOM.
Advanced control
trainNeighbourhood :: (Pattern p, FiniteGrid (gm p), GridMap gm p, Num (Metric p), Ord k, k ~ Index (gm p), k ~ Index (BaseGrid gm p), Fractional (Metric p)) => DSOM gm t p -> k -> p -> DSOM gm k pSource
Trains the specified node and the neighbourood around it to better
match a target.
Most users should use train
, which automatically determines
the BMU and trains it and its neighbourhood.