Classification with Costly Features in Hierarchical Deep Sets
This work addresses the limitation of existing costly feature classification methods that cannot handle hierarchical data, which is common in real-world applications like web security, though it is incremental as it builds on prior deep reinforcement learning approaches.
The paper tackles the problem of classification with costly features for hierarchical data, extending a deep reinforcement learning method with hierarchical deep sets and hierarchical softmax to process complex structures like XML or JSON, and shows superior performance in experiments on seven datasets and a real-world malicious web domain classification task.
Classification with Costly Features (CwCF) is a classification problem that includes the cost of features in the optimization criteria. Individually for each sample, its features are sequentially acquired to maximize accuracy while minimizing the acquired features' cost. However, existing approaches can only process data that can be expressed as vectors of fixed length. In real life, the data often possesses rich and complex structure, which can be more precisely described with formats such as XML or JSON. The data is hierarchical and often contains nested lists of objects. In this work, we extend an existing deep reinforcement learning-based algorithm with hierarchical deep sets and hierarchical softmax, so that it can directly process this data. The extended method has greater control over which features it can acquire and, in experiments with seven datasets, we show that this leads to superior performance. To showcase the real usage of the new method, we apply it to a real-life problem of classifying malicious web domains, using an online service.