“…Broadly speaking, random projections offer a universal and flexible approach to complex statistical problems. They are a particularly useful tool in large‐scale settings, such as high‐dimensional classification (Cannings & Samworth, 2017; Durrant & Kabán, 2013, 2015), clustering (Dasgupta, 1999; Fern & Brodley, 2003; Heckel, Tschannen, & Bölcskei, 2017), precision matrix estimation (Marzetta, Tucci, & Simon, 2011), regression (Ahfock, Astle, & Richardson, 2017; Dobriban & Liu, 2019; Heinze, McWiliams, & Meinshausen, 2016; Klanke, Vijayakumar, & Schaal, 2008; McWilliams, Heinze, Meinshausen, Krummenacher, & Vanchinathan, 2014; Mukhopadhyay & Dunson, 2019; Slawski, 2018; Thanei, Heinze, & Meinshausen, 2017; Thanei, Meinshausen, & Shah, 2018), sparse principal component analysis (Gataric, Wang, & Samworth, 2019), hypothesis testing (Lopes, Jacob, & Wainwright, 2011; Shi, Lu, & Song, 2019), correlation estimation (Grellmann et al, 2016), dimension reduction (Bingham & Mannilla, 2001; Omidiran & Wainwright, 2010; Reeve, Mu, & Brown, 2018), and matrix decomposition (Halko, Martinsson, & Tropp, 2011).…”