Learning With Feature Evolvable Streams

By Bo-Jian Hou, Lijun Zhang, Zhi-Hua Zhou. 2017

84 citations recorded by the source corpus.

Learning with streaming data has attracted much attention during the past few years. Though most studies consider data stream with fixed features, in real practice the features may be evolvable. For example, features of data gathered by limited-lifespan sensors will change when these sensors are substituted by new ones. In this article, we propose a novel learning paradigm: Feature Evolvable Streaming Learning where old features would vanish and new features would occur. Rather than relying on only the current features, we attempt to recover the vanished features and exploit it to improve performance. Specifically, we learn a mapping from the overlapping period to recover old features and then we learn two models from the recovered features and the current features, respectively. To benefit from the recovered features, we develop two ensemble methods. In the first method, we combine the predictions from two models and theoretically show that with the assistance of old features, the performance on new features can be improved and we provide a tighter bound when the loss function is exponentially concave. In the second approach, we dynamically select the best single prediction and establish a better performance guarantee when the best model switches. Experiments on both synthetic and real data validate the effectiveness of our proposal.

Subfields: Artificial Intelligence, Signal Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Xinchao Wang
2. Mingli Song
3. Peng Zhao
4. Jingwen Ye
5. Jie Song
6. Yuan Jiang
7. Zhi-Hua Zhou
8. Xingyi Yang

In Signal Processing

1. Peng Zhao
2. Zhi-Hua Zhou
3. Bo-Jian Hou
4. Mingli Song
5. Yuan Jiang
6. Jingwen Ye
7. Jie Song
8. Xinchao Wang

Citations in the corpus

• Learning with Recoverable Forgetting (2022, positive): “In addition to the above works that tend to forbid the catastrophic forgetting of the old tasks, some researchers [19,64,18,17] pay more attention on the decremental cases where some features may vanish while feature evolving.”
• Storage Fit Learning with Feature Evolvable Streams (2020, methodology): “We focus on FESL [HZZ17a], and other feature evolvable learning methods based on online learning technique can also adapt to our framework.”
• Learning with Feature and Distribution Evolvable Streams (2020, methodology): “We perform the EDM algorithm and the FESL algorithm (Hou et al., 2017) on these three synthetic tasks, where both algorithms return the optimized discrepancy.”