[et_pb_section fb_built=”1″ _builder_version=”4.19.4″ _module_preset=”default” global_colors_info=”{}”][et_pb_row _builder_version=”4.19.4″ _module_preset=”default” global_colors_info=”{}”][et_pb_column type=”4_4″ _builder_version=”4.19.4″ _module_preset=”default” global_colors_info=”{}”][et_pb_text _builder_version=”4.19.4″ _module_preset=”default” custom_padding=”||20px|||” global_colors_info=”{}”]<\/p>\n
This article published in Nature Machine Intelligence<\/em> explains how deep-learning classification algorithms can learn non-medical and irrelevant information by making “shortcuts”, and then replicate them in an external validation base.<\/p>\n This article highlights the importance of the quality of the data selected for AI learning to reduce “shortcuts”. Even so, the possibility of irrelevant learning is not entirely absent, so it seems necessary to focus on explainable AI approaches to overcome this bias. <\/p>\n [\/et_pb_text][et_pb_button button_url=”https:\/\/www.nature.com\/articles\/s42256-021-00338-7″ url_new_window=”on” button_text=”View article” _builder_version=”4.19.4″ _module_preset=”default” custom_padding=”8px||8px||true|false” global_colors_info=”{}”][\/et_pb_button][\/et_pb_column][\/et_pb_row][\/et_pb_section]<\/p>\n","protected":false},"excerpt":{"rendered":" This article explains how deep learning classification algorithms can learn non-medical, irrelevant information, then replicate it in an external validation base.<\/p>\n","protected":false},"author":1,"featured_media":2832,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_et_pb_use_builder":"on","_et_pb_old_content":"","_et_gb_content_width":"","footnotes":""},"categories":[48],"tags":[],"class_list":["post-2999","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-news-ai-in-healthcare"],"yoast_head":"\n