Recent advances in molecular biotechnologies demand fast and innovative algorithms to analyze them effectively. One promising proximal-type algorithm is iterative hard thresholding (IHT). It is especially well suited to the analysis of modern high-dimensional datasets common in genomics, and was recently implemented in Julia as IHT.jl for both numeric and genomic data. Compared to related packages, the current IHT implementation excels in aspects such as memory management and model selection, but lacks important analysis features such as group selection relevant to genomic analysis. Hence, IHT.jl is ripe for improvement.



Benjamin Chu


  • Mike Innes
  • Kevin Keys
  • Christopher Rackauckas