图 书 馆
一次学习的两种架构Two architectures for one-shot learning |
|
| 课程网址: | http://videolectures.net/nipsworkshops2013_tenenbaum_learning/ |
| 主讲教师: | Joshua B. Tenenbaum |
| 开课单位: | 麻省理工学院 |
| 开课时间: | 2014-10-06 |
| 课程语种: | 英语 |
| 中文简介: |  人们可以从一个示例中几乎完美地学习新的视觉概念，但机器学习算法通常需要数百个示例才能执行类似的操作。人类还可以以比传统机器学习系统更丰富的方式使用他们学到的概念，将对象解析为多个部分，生成新实例，识别概念的抽象类型，甚至想象给定类型的新概念。我将讨论我们探索的两种计算方法，旨在捕捉这些人类学习能力。第一个（与 Ruslan Salakhutdinov 和 Antonio Torralba 合作）是分层深度或“HD”架构，它在深度 Boltzmann 机器特征提取器之上学习分层非参数贝叶斯模型（特别是分层 Dirichlet 过程主题模型）。这种方法非常通用，但我认为它的结构不足以捕捉人类在现实世界领域中学习的概念知识。然后，我将展示一个新架构（与 Brendan Lake 和 Ruslan Salakhutdinov），它将概念表示为简单的程序，体现层次结构、组合性和因果关系的基本原则，并构建在贝叶斯标准下最好地解释观察到的示例的程序。这种方法在选择要学习的程序形式时需要更多领域特定的工程，但它仍然非常灵活。在具有挑战性的一次性分类任务中，贝叶斯程序学习器是第一个达到人类水平性能的，大大优于一系列其他方法，包括深度玻尔兹曼机器和我们的 HD 模型。我还将举例说明几个“视觉图灵测试”实验，探讨程序学习模型更具创造性的解析、泛化和生成能力，并表明在许多情况下它与人类的表现没有区别。 |
| 课程简介: | People can learn a new visual concept almost perfectly from just a single example, yet machine learning algorithms typically require hundreds of examples to perform similarly. Humans can also use their learned concepts in richer ways than conventional machine learning systems, to parse objects into parts, generate new instances,recognize abstract types of concepts and even imagine novel concepts of a given type. I will discuss two computational approaches we have explored that aim to capture these human learning abilities. The first (with Ruslan Salakhutdinov and Antonio Torralba) is the hierarchical-deep or "HD" architecture, which learns a hierarchical nonparametric Bayesian model (specifically, a hierarchical Dirichlet process topic model) on top of a deep Boltzmann machine feature extractor. This approach is appealingly general, but I will argue it is insufficiently structured to capture the conceptual knowledge humans learn in real-world domains. I will then present a new architecture (with Brendan Lake and Ruslan Salakhutdinov) that represents concepts as simple programs, embodying basic principles of hierarchy, compositionality and causality, and constructs programs that best explain observed examples under a Bayesian criterion. This approach requires more domain-specific engineering in choosing the form of the programs to be learned, but it is still very flexible. On a challenging one-shot classification task, the Bayesian program learner is the first to achieve human-level performance, substantially outperforming a range of other approaches including both deep Boltzmann machines and our HD models. I will also illustrate several "visual Turing test" experiments probing the program learning model's more creative parsing, generalization and generation abilities, and show that in many cases it is indistinguishable from the performance of humans. |
| 关 键 词: | 机器学习; 视觉图灵测试; 特征提取器 |
| 课程来源: | 视频讲座网 |
| 数据采集: | 2021-06-17：liyy |
| 最后编审: | 2021-06-17：liyy |
| 阅读次数: | 24 |