| --- | |
| response: | |
| apps: | |
| - appid: 228980 | |
| app: Steamworks Common Redistributables | |
| app_type: tool | |
| available_on_platform: true | |
| - appid: 250820 | |
| app: SteamVR | |
| app_type: tool |
Python implementaion of HAM as deseibed in Functional Annotation of Genes Using Hierarchical Text Categorization
Most researchers evaluate hierarchical classification systems based on standard “flat” measures: accuracy/error and precision/recall. However, these measures are not suitable for hierarchical categorization since they do not differentiate among different kinds of misclassification errors. A widely-used hierarchical measure based on the notion of distance overcomes this problem. However, it has some drawbacks. First, it is not easily extendable to DAG hierarchies (where multiple paths between two categories can exist) and multi-label tasks. Second, it does not change with depth. Misclassification into a sibling category of a top