With the booming of social networks, such as Facebook and Flickr, the candidate labels of an instance can be numerous. Hence, traditional multi-label learning algorithms are out of capability to handle a large quantity of labels for the unaffordable time complexity. To alleviate this problem, label space dimension reduction (LSDR) is proposed by transforming the original label space into a lower dimensional one. Inspired by the effectiveness of coupled dictionary learning (CDL) in dealing with cross-modal data, in this paper, we proposed a novel algorithm named Coupled Dictionary Learning for Multi-label Embedding (ML-CDL) to track the problem of LSDR. We novelly treat feature and label as coupled domains. Then CDL is utilized to generate the low-dimensional latent space that leverages the information between feature and label spaces. In particular, the sparse representation coefficients embody the properties of interpretability, discriminability and sparsity. Experimental results on benchmark datasets demonstrate the effectiveness of our algorithm.

MOTIVATION: Automatic error correction of high-throughput sequencing data can have a dramatic impact on the amount of usable base pairs and their quality. It has been shown that the performance of tasks such as de novo genome assembly and SNP calling can be dramatically improved after read error correction. While a large number of methods specialized for correcting substitution errors as found in Illumina data exist, few methods for the correction of indel errors, common to technologies like 454 or Ion Torrent, have been proposed. RESULTS: We present Fiona, a new stand-alone read error-correction method. Fiona provides a new statistical approach for sequencing error detection and optimal error correction and estimates its parameters automatically. Fiona is able to correct substitution, insertion and deletion errors and can be applied to any sequencing technology. It uses an efficient implementation of the partial suffix array to detect read overlaps with different seed lengths in parallel. We tested Fiona on several real datasets from a variety of organisms with different read lengths and compared its performance with state-of-the-art methods. Fiona shows a constantly higher correction accuracy over a broad range of datasets from 454 and Ion Torrent sequencers, without compromise in speed. CONCLUSION: Fiona is an accurate parameter-free read error-correction method that can be run on inexpensive hardware and can make use of multicore parallelization whenever available. Fiona was implemented using the SeqAn library for sequence analysis and is publicly available for download at http://www.seqan.de/projects/fiona. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.