A Combined Selection of Fragmentation and Allocation Schemes in Parallel Data Warehouses

Viana, J. M. S., Faria, V. R., Fonseca e Silva, F. and Vilela de Resende, M. D. 2012. Combined selection of progeny in crop breeding using best linear unbiased prediction. Can. J. Plant Sci. 92: 553–562. Combined selection is an important strategy in crop breeding. As the classical index does not consider pedigree information, the objective of this study was to evaluate the efficiency of the best linear unbiased prediction (BLUP) methodology for combined selection of progeny. We analyzed expansion volume (EV) and grain yield of parents and inbred and non-inbred progeny from the popcorn population Viçosa. The BLUP analyses, single-trait and of the same character measured in parents and progeny (combined parent-family) were performed using the ASReml software. Because the experiments were balanced, the estimates of the additive variance from the BLUP and least squares analyses were generally equivalent. The accuracies of the BLUP analyses do not clearly establish the superior technique. The accuracy of the classical index tended to be higher than that obtained from BLUP analyses. There was equivalence between BLUP and least squares analyses relative to half-sib and inbred progeny selection, and superiority of the combined parent-family BLUP index for full-sib selection. The BLUP analyses also differed from the least squares analysis on the coincidence of selected parents. The populations obtained by selection based on BLUP of breeding values presented a lower effective size.

Download Full-text

Ensemble and Deep Learning for Language-Independent Automatic Selection of Parallel Data

Algorithms ◽

10.3390/a12010026 ◽

2019 ◽

Vol 12 (1) ◽

pp. 26 ◽

Cited By ~ 2

Author(s):

Despoina Mouratidis ◽

Katia Kermanidis

Keyword(s):

Deep Learning ◽

Machine Translation ◽

Research Field ◽

Attribute Selection ◽

Parallel Corpora ◽

Automatic Selection ◽

Parallel Data ◽

First Time ◽

Learning Architectures ◽

Selection Of

Machine translation is used in many applications in everyday life. Due to the increase of translated documents that need to be organized as useful or not (for building a translation model), the automated categorization of texts (classification), is a popular research field of machine learning. This kind of information can be quite helpful for machine translation. Our parallel corpora (English-Greek and English-Italian) are based on educational data, which are quite difficult to translate. We apply two state of the art architectures, Random Forest (RF) and Deeplearnig4j (DL4J), to our data (which constitute three translation outputs). To our knowledge, this is the first time that deep learning architectures are applied to the automatic selection of parallel data. We also propose new string-based features that seem to be effective for the classifier, and we investigate whether an attribute selection method could be used for better classification accuracy. Experimental results indicate an increase of up to 4% (compared to our previous work) using RF and rather satisfactory results using DL4J.

Download Full-text

Porantim-Opt: Optimizing the Combined Selection of Model-Based Testing Techniques

2011 IEEE Fourth International Conference on Software Testing, Verification and Validation Workshops ◽

10.1109/icstw.2011.33 ◽

2011 ◽

Cited By ~ 3

Author(s):

Arilo Claudio Dias Neto ◽

Rosiane de Freitas Rodrigues ◽

Guilherme Horta Travassos

Keyword(s):

Model Based ◽

Model Based Testing ◽

Testing Techniques ◽

Combined Selection ◽

Selection Of

Download Full-text

SELECTION OF SIRES ORIGINATING FROM A NUCLEUS BREEDING UNIT FOR USE IN A COMMERCIAL DAIRY POPULATION

Canadian Journal of Animal Science ◽

10.4141/cjas89-071 ◽

1989 ◽

Vol 69 (3) ◽

pp. 595-604 ◽

Cited By ~ 2

Author(s):

GEORG TEEPKER ◽

DAVID S. KELLER

Keyword(s):

Embryo Transfer ◽

Pedigree Information ◽

Sources Of Information ◽

Production Environment ◽

Genetic Progress ◽

Genotype By Environment Interactions ◽

Genotype By Environment ◽

Commercial Population ◽

Combined Selection ◽

Selection Of

Bulls that originate from breeding schemes with extensive use of multiple ovulation and embryo transfer (MOET) can be selected on pedigree information, the performance of their contemporary female full- and half-sibs or based on a progeny test. Combined selection across groups of bulls with different sources of information is shown to be superior to selection of progeny tested bulls alone. The magnitude of the superiority for a given selection rate is determined by the differences in genetic merit between the groups and ranges from 5 to 65%. Depending on the annual genetic progress the majority of the bulls are selected either on pedigree or progeny information. The proportion of sib-tested bulls ranges from 9 to 23%. When breeding activities are concentrated in nucleus herds, possible genotype by environment interactions between performance in the test herd and in the commercial population have to be considered. Genotype by environment interactions increase the proportion of progeny tested sires that will be selected. However, the proportion of progeny tested sires used is reduced by genotype by environment interactions among the groups of a heterogenous commercial production environment. Key words: Embryo transfer, sire selection, genotype-by-environment interactions

Download Full-text