Automating t-SNE Parameterization with Prototype-Based Learning of Manifold Connectivity
Automating t-SNE Parameterization with Prototype-Based Learning of Manifold Connectivity
复制标题
通过基于原型的流形连接学习实现 t-SNE 参数化自动化
DOI：
10.1016/j.neucom.2022.07.009
复制
发表时间：
2022
期刊：
影响因子：
6
通讯作者：
Merényi, E.
中科院分区：
文献类型：
--
作者：
Taylor, J;Merényi, E.
我们利用通过基于神经原型的学习来自动化t-SNE参数化的数据流形的拓扑信息。这些信息包含在神经原型的CONN（连通性）相似性中，它对数据流形内各个点的拓扑连通性的强度（弱点）进行分级。CONN提出了一个数据驱动的规范的本地化版本（不同的歧管）的t-SNE的困惑参数，反过来，定义了高维similaritiesP的t-SNE试图保存。我们进一步imbueP与CONN的分级相似性，以减少流形的拓扑结构和其嵌入表示之间的不匹配。实验表明，这些改进，统称为SNE，能够产生有意义的和值得信赖的低维嵌入，而不需要在（即，网格搜索）t-SNE的困惑空间。数据驱动的t-SNE参数化提高了我们的信心，即嵌入中出现的任何结构都是有效的，而不仅仅是虚假参数化的伪像。
We harness topological information about a data manifold revealed through neural prototype-based learning to automate t-SNE parameterization. This information is contained in the CONN (CONNectivity) similarity of neural prototypes, which grades the strength (weakness) of topological connectivity at various points within a data manifold. CONN suggests a data-driven specification of localized versions (varying across the manifold) of t-SNE’s perplexity parameter which, in turn, defines the high-dimensional similaritiesPthat t-SNE attempts to preserve. We further imbuePwith CONN’s graded similarity to reduce mismatch between the topology of the manifold and its embedded representation. Experiments show these improvements, collectively calledCONNt-SNE, are capable of producing meaningful and trustworthy low-dimensional embeddings without the need to heuristically optimize over (i.e., grid search) t-SNE’s perplexity space. Data-driven t-SNE parameterization improves our confidence that any structure appearing in the embeddings is valid and not merely an artifact of spurious parameterization.