Knowledge-Based Fast Evaluation for Evolutionary Learning

Giráldez, Raúl; Aguilar–Ruiz, Jesús S.; Riquelme, José C.

doi:10.1109/tsmcc.2004.841904

Cited by 26 publications

(19 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…After all, fixed finite algorithmic randomization algorithms would only qualify as ''weak learners'' over arbitrary domains. Only knowledge-based learning can be said to be truly strong [9] and that knowledge itself must be randomized -allowing for the creative induction of the same.…”

Section: Prefacementioning

confidence: 99%

See 1 more Smart Citation

On randomization and discovery

Rubin

2007

Information Sciences

View full text Add to dashboard Cite

Section: Prefacementioning

confidence: 99%

“…The complexity of a series of digits is the number of bits that must be put into a computing machine in order to obtain the original series as output. As a consequence, all effective learning methods must be knowledge-based, or strong, if they are not to be trivial [9]. Proposition 3.1.…”

Section: Unsolvability Of the Randomization Problemmentioning

confidence: 99%

On randomization and discovery

Rubin

2007

Information Sciences

View full text Add to dashboard Cite

“…In the same manner, the memory footprint required to store the population of rules also grows as O( r · c). Recent studies have proposed to used index structures to speedup the matching process [4]. Tree-based indexing may reduce the required time for matching and instance to O( r · log( c)).…”

Section: Rule Sets As Listsmentioning

confidence: 99%

Fast rule matching for learning classifier systems via vector instructions

Llorà

Sastry

2006

Proceedings of the 8th Annual Conference on Genetic and Evolutionary Computation

View full text Add to dashboard Cite

Over the last ten years XCS has become the standard for Michigan-style learning classifier systems (LCS). Since the initial CS-1 work conceived by Holland, classifiers (rules) have widely used a ternary condition alphabet {0,1,#} for binary input problems. Most of the freely available implementations of this ternary alphabet in XCS rely on character-based encodings-easy to implement, not memory efficient, and expensive to compute. Profiling of freely available XCS implementations shows that most of their execution time is spent determining whether a rule is match or not, posing a serious threat to XCS scalability. In the last decade, multimedia and scientific applications have pushed CPU manufactures to include native support for vector instruction sets. This paper presents how to implement efficient condition encoding and fast rule matching strategies using vector instructions. The paper elaborates on Altivec (PowerPC G4, G5) and SSE2 (Intel P4/Xeon and AMD Opteron) instruction sets producing speedups of XCS matching process beyond ninety times. Moreover, such a vectorized matching code will allow to easily scale beyond tens of thousands of conditions in a reasonable time. The proposed fast matching scheme also fits in any other LCS other than XCS.

show abstract

“…[22] contains a taxonomy of such methods. A different approach consists in "precomputing" the possible classifications of the system [25]. The authors exploit the fact that they deal with either nominal or discretized attributes to generate tree structure to efficiently know which examples belong to each value of each attribute (the one corresponding to the rule being evaluated).…”

Section: Related Workmentioning

confidence: 99%

“…In recent years, many techniques have been developed to reduce the computational cost of EL methods by using mechanisms such as windowing techniques [1], precomputing the possible matches of the individuals [25] or using the vectorial instructions (SSE, Altivec) available in modern day computers [31].…”

Section: Introductionmentioning

confidence: 99%

Improving the scalability of rule-based evolutionary learning

2008

View full text Add to dashboard Cite

Evolutionary learning techniques are comparable in accuracy with other learning methods such as Bayesian Learning, SVM, etc. These techniques often produce more interpretable knowledge than, e.g. SVM; however, efficiency is a significant drawback. This paper presents a new representation motivated by our observations that Bioinformatics and Systems Biology often give rise to very large-scale datasets that are noisy, ambiguous and usually described by a large number of attributes. The crucial observation is that, in the most successful rules obtained for such datasets, only a few key attributes (from the large number of available ones) are expressed in a rule, hence automatically discovering these few key attributes and only keeping track of them contributes to a substantial speed up by avoiding useless match operations with irrelevant attributes. Thus, in effective terms this procedure is performing a fine-grained feature selection at a rule-wise level, as the key attributes may be different for each learned rule. The representation we propose has been tested within the BioHEL machine learning system, and the experiments performed show that not only the representation has competent learning performance, but that it also manages to reduce considerably the system run-time. That is, the proposed representation is up to 2-3 times faster than state-of-the-art evolutionary learning representations designed specifically for efficiency purposes

show abstract

Knowledge-Based Fast Evaluation for Evolutionary Learning

Cited by 26 publications

References 25 publications

On randomization and discovery

On randomization and discovery

Fast rule matching for learning classifier systems via vector instructions

Improving the scalability of rule-based evolutionary learning

Contact Info

Product

Resources

About