seq2feature_mds_large extracts MDS features from a large number of
response processes. The algorithm proposed in Paradis (2018) is implemented with minor
variations to perform MDS. The algorithm first selects a relatively small subset of
response processes to perform the classical MDS. Then the coordinate of each of the
other response processes are obtained by minimizing the loss function related to the target
response processes and the those in the subset through BFGS.
seq2feature_mds_large(seqs, K, dist_type = "oss_action", subset_size,
subset_method = "random", n_cand = 10, pca = TRUE, seed = 12345,
L_set = 1:3)an object of class "proc"
the number of features to be extracted.
a character string specifies the dissimilarity measure for two response processes. See 'Details'.
the size of the subset on which classical MDS is performed.
a character string specifying the method for choosing the subset.
It must be one of "random", "sample_avgmax",
"sample_minmax", "full_avgmax", and "full_minmax".
The size of the candidate set when selecting the subset. It is only used when
subset_method is "sample_avgmax" or "sample_minmax".
logical. If TRUE (default), the principal components of the
extracted features are returned.
random seed.
length of ngrams considered
seq2feature_mds_large returns an \(n \times K\) matrix of extracted
features.
Paradis, E. (2018). Multidimensional Scaling with Very Large Datasets. Journal of Computational and Graphical Statistics, 27, 935--939.
Other feature extraction methods: aseq2feature_seq2seq,
atseq2feature_seq2seq,
seq2feature_mds,
seq2feature_seq2seq,
tseq2feature_seq2seq