Name
Abstract | ||
|---|---|---|
We human beings are often careless, forgetful and ignorant. We often miss good timing for capturing chances or avoiding risks. Computational awareness (CA) can help us to be more aware. An aware system can distinguish novel events from normal ones, inform us when the novel events are detected, and tell us the correct reaction. In many cases, we are interested in knowing reasons why certain situations are novel, and why certain reactions are necessary under these situations. The ability of providing understandable reasons can make the system more trustable. Designing interpretable aware systems should be an important goal of CA researches. In this article, we provide a method for translating an aware system into an expert system that in turn can be used to provide reasons for making decision. As a case study, we show the process for interpreting a learned 3-valuved logic multilayer perceptron. The proposed method should be useful for achieving the goal of CA. |
Year | DOI | Venue |
|---|---|---|
2016 | 10.1109/ICMLC.2016.7873003 | 2016 International Conference on Machine Learning and Cybernetics (ICMLC) |
Keywords | Field | DocType |
Computational awareness,Aware systems,3-valued logic,Multilayer perceptron,Expert system | Computer science,Expert system,Multilayer perceptron,Artificial intelligence,Cognition,Machine learning,Cybernetics | Conference |
Volume | ISBN | Citations |
2 | 978-1-5090-0391-4 | 3 |
PageRank | References | Authors |
0.69 | 0 | 1 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Qiangfu Zhao | 1 | 214 | 62.36 |