Different scaling of linear models and deep learning in UKBiobank brain images versus machine-learning datasets

• Published in: Nature communications Vol. 11; no. 1; pp. 4238 - 15
• Main Authors: Schulz, Marc-Andre, Yeo, B. T. Thomas, Vogelstein, Joshua T., Mourao-Miranada, Janaina, Kather, Jakob N., Kording, Konrad, Richards, Blake, Bzdok, Danilo
• Format: Journal Article
• Language: English
• Published: London Nature Publishing Group UK 25.08.2020; Nature Publishing Group; Nature Portfolio
• Subjects: 49; 59/36; 59/57; 631/378/116/2394; 706/648/697/129/2043; Algorithms; Biological Specimen Banks; Brain; Brain - diagnostic imaging; Datasets; Deep Learning; Depth profiling; Humanities and Social Sciences; Humans; Kernels; Learning algorithms; Linear Models; Machine Learning; multidisciplinary; Neuroimaging - methods; Phenotype; Phenotypes; Predictions; Sample Size; Science; Science (multidisciplinary); Structure-function relationships; United Kingdom; United Kingdom

Recently, deep learning has unlocked unprecedented success in various domains, especially using images, text, and speech. However, deep learning is only beneficial if the data have nonlinear relationships and if they are exploitable at available sample sizes. We systematically profiled the performan...