random_projection
Random projection reducer for continuous datasets. The library
implements the dimension_reducer_protocol defined in the
dimension_reduction_protocols library and learns a seeded dense
Rademacher projection matrix using the portable fast_random
pseudo-random generator after centering the training data, optionally
standardizing continuous attributes, and sampling entries in
{-$1/sqrt(k)$, +$1/sqrt(k)$} where $k$ is the requested reduced
dimensionality.
API documentation
Open the ../../apis/library_index.html#random_projection link in a web browser.
Loading
To load this library, load the loader.lgt file:
| ?- logtalk_load(random_projection(loader)).
Testing
To test this library predicates, load the tester.lgt file:
| ?- logtalk_load(random_projection(tester)).
Features
Continuous Datasets: Accepts datasets containing only continuous attributes. Missing or nonnumeric values are rejected.
Centering and Optional Scaling: Centers all attributes and optionally standardizes them before projection.
Portable Seeded Sampling: Uses
fast_random(xoshiro128pp)so learned projection matrices are portable and reproducible.Projection API: Transforms a new instance into a list of
component_N-Valuepairs.Model Export: Learned reducers can be exported as predicate clauses or written to a file.
Options
The learn/3 predicate accepts the following options:
n_components/1: Number of random projection components to sample. Requests that exceed the number of features raisedomain_error(component_count, Requested-Maximum). The default is2.feature_scaling/1: Whether to standardize continuous attributes before projection. Options:true(default) orfalse.random_seed/1: Positive integer used to seed the portable pseudo-random generator before sampling the projection matrix. The default is1357911.
Usage
The following examples use the sample datasets shipped with the
dimension_reduction_protocols library:
| ?- logtalk_load(dimension_reduction_protocols('test_datasets/correlated_plane')),
logtalk_load(dimension_reduction_protocols('test_datasets/high_dimensional_measurements')).
Learning a reducer
| ?- random_projection::learn(correlated_plane, DimensionReducer).
| ?- random_projection::learn(correlated_plane, DimensionReducer, [n_components(1), feature_scaling(false), random_seed(17)]).
Transforming new instances
| ?- random_projection::learn(high_dimensional_measurements, DimensionReducer, [random_seed(11)]),
random_projection::transform(DimensionReducer, [f1-0.9, f2-1.1, f3-1.0, f4-2.0, f5-2.2, f6-2.1], ReducedInstance).
| ?- random_projection::learn(correlated_plane, DimensionReducer, [n_components(1), random_seed(19)]),
random_projection::transform(DimensionReducer, [x-1.0, y-2.0, z-3.0], ReducedInstance).
Exporting and reusing the reducer
| ?- random_projection::learn(correlated_plane, DimensionReducer, [n_components(1), random_seed(29)]),
random_projection::export_to_file(correlated_plane, DimensionReducer, reducer, 'random_projection_reducer.pl').
| ?- logtalk_load('random_projection_reducer.pl'),
reducer(Reducer),
random_projection::transform(Reducer, [x-1.0, y-2.0, z-3.0], ReducedInstance).
Dimension reducer representation
The learned dimension reducer is represented by a compound term with the functor chosen by the implementation and arity 3. For example:
random_projection_reducer(Encoders, Components, Diagnostics)
Where:
Encoders: List of continuous attribute encoders storing attribute name, mean, and scale.Components: List of sampled projection vectors in component order.Diagnostics: Learned reducer metadata including the effective training options and reproducibility details.
When exported using export_to_clauses/4 or export_to_file/4,
this reducer term is serialized directly as the single argument of the
generated predicate clause so that the exported model can be loaded and
reused as-is.
References
Johnson, W. B. and Lindenstrauss, J. (1984) - “Extensions of Lipschitz mappings into a Hilbert space”.
Achlioptas, D. (2003) - “Database-friendly random projections: Johnson-Lindenstrauss with binary coins”.