#2018 ML course project in IEEE Honor Class. ## 基于分子显微图像的蛋白质亚细胞定位 My View: 这个project label不准确...... sota只能做到50%-60% dataset description 人类蛋白图库（Human Protein Atlas, HPA）是一个公开免费的生物图像数据库，目前的Version 16.1中储存有16998个蛋白质在人类健康和癌症组织以及细胞中的上百万幅显微图像。 difficulty * 图片尺寸大： 3000*3000 * 图片有用信息少、噪声高 * Multi-label * Multi-instance related work * Image based: Xu, Y., et al., Bioimaging-based detection of mislocalized proteins in human cancers by semi-supervised learning. Bioinformatics, 2014. 31(7): p. 1111-1119. * Image based: Xu, Y.Y., et al., An image-based multi-label human protein subcellular localization predictor (iLocator) reveals protein mislocalizations in cancer tissues. Bioinformatics, 2013. 29(16): p. 2032-2040 * Sequence based: Almagro, J.A., et al., DeepLoc: prediction of protein subcellular localization using deep learning. Bioinformatics, 2017.