Machine learning for classification?

[a-ludi]

Machine learning for classification?

[a-ludi]

Furthermore, theses approaches might be used to reduce the candidate set, ie. filter those reads that likely do not belong to this locus. Additional information such as the pairwise alignment of the candidates to each other or the actual sequence should be used.

[a-ludi]

One point of critique of GMcloser is that it has been trained on a set of rice data only – even though they do not use machine learning they need to estimate some parameters. Machine learning approaches require a training set and that introduces a number of problems. Some of them conflicts with our goal "keep it simple". Consequently, we discard this idea.