Abstract | ||
---|---|---|
There are many classifiers that treat entities to be classified as points in a high-dimensional vector space and then compute a separator S between entities in class +1 from those in class -1. However, such classifiers are usually very hard to explain in plain English to domain experts. We propose Metric Logic Programs (MLPs) which are a fragment of constraint logic programs as a new paradigm for explaining S. We present multiple measures of quality of an MLP and define the problem of finding an MLP-Explanation of S and show that it - and various related problems - are NP-hard. We present the MLP Extract algorithm to extract MLP explanations for S. We show that while our algorithms provide more succinct, simpler, and higher fidelity explanations than association rules that are less expressive, our algorithms do require additional run-time. |
Year | DOI | Venue |
---|---|---|
2016 | 10.1007/978-3-319-45856-4_14 | Lecture Notes in Artificial Intelligence |
Field | DocType | Volume |
Signature (logic),Computational logic,Fidelity,Vector space,Formal language,Computer science,Support vector machine,Theoretical computer science,Association rule learning,Artificial intelligence,Machine learning,Ontology language | Conference | 9858 |
ISSN | Citations | PageRank |
0302-9743 | 0 | 0.34 |
References | Authors | |
8 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Srijan Kumar | 1 | 326 | 24.97 |
Edoardo Serra | 2 | 24 | 4.03 |
Francesca Spezzano | 3 | 80 | 19.08 |
V. S. Subrahmanian | 4 | 6864 | 1053.38 |