Large scale machine learning using GPUs

使用 GPU 进行大规模机器学习

• 批准号: 406084-2011
• 负责人: Hinton, Geoffrey
• 金额: $ 4.37万
• 依托单位: University of Toronto
• 依托单位国家: 加拿大
• 项目类别: Research Tools and Instruments - Category 1 (<$150,000)
• 财政年份: 2010
• 资助国家: 加拿大
• 起止时间: 2010-01-01 至 2011-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=472384
• 关键词: Large; scale; machine; learning; using

Recently discovered machine learning algorithms are very good at learning multiple layers of features. These features are very good for a variety of important tasks such as recognizing three-dimensional objects in images, automatically extracting road networks from cluttered aerial images of urban areas, recognizing words in speech, and retrieving images that are semantically similar to a query image. Related learning algorithms are very good at recommending products based on a user's rating of other products.

最近发现的机器学习算法非常擅长学习多层特征。这些特征对于识别图像中的三维物体、从城市地区杂乱的航空图像中自动提取道路网络、识别语音中的单词以及检索与查询图像在语义上相似的图像等各种重要任务都非常有用。相关的学习算法非常擅长根据用户对其他产品的评价来推荐产品。