- author = "Hugh Leather and Edwin Bonilla and Michael O'Boyle",
- title = "Automatic Feature Generation for Machine Learning Based Optimizing Compilation",
- booktitle = "2009. International Symposium on Code Generation and Optimization",
- year = "2009",
- publisher = "IEEE",
- address = "Seattle, WA, USA",
- month = "22-25 " # mar,
- pages = "81--91",
- keywords = "genetic algorithms, genetic programming, grammatical evolution, SBSE, Pentium 6, automatic feature generation, compilation, compiler writer, feature generation technique, grammar, loop unrolling, machine learning, predictive modeling, grammars, learning (artificial intelligence), program compilers",
- isbn13 = "978-0-7695-3576-0",
-
DOI = "
doi:10.1109/CGO.2009.21",
- abstract = "Recent work has shown that machine learning can automate and in some cases outperform hand crafted compiler optimizations. Central to such an approach is that machine learning techniques typically rely upon summaries or features of the program. The quality of these features is critical to the accuracy of the resulting machine learned algorithm; no machine learning method will work well with poorly chosen features. However, due to the size and complexity of programs, theoretically there are an infinite number of potential features to choose from. The compiler writer now has to expend effort in choosing the best features from this space. This paper develops a novel mechanism to automatically find those features which most improve the quality of the machine learned heuristic. The feature space is described by a grammar and is then searched with genetic programming and predictive modeling. We apply this technique to loop unrolling in GCC 4.3.1 and evaluate our approach on a Pentium 6. On a benchmark suite of 57 programs, GCC's hard-coded heuristic achieves only 3percent of the maximum performance available, while a state of the art machine learning approach with hand-coded features obtains 59percent. Our feature generation technique is able to achieve 76percent of the maximum available speedup, outperforming existing approaches.",
-
notes = "p85 'Our search technique is a hybrid between
Grammatical Evolution [12] and Genetic Programming
[13].'
Also known as \cite{4907653}",
