USE OF TECHNICAL OF K-MEANS CLUSTER TO FORMATION SUB-GROUPS FOR THE PHYSICAL TRAINING IN SOCCER TEAMFederal University of Parana - Center for the Study of Physical Performance - Curitiba-PR BrazilVitor B. Nascimento, Luiz F. Novack, Fabiano , Luis F. Carignano, Raul Osiecki
E-mail: vitorbertolinascimento@yahoo.com.br
Introduction: A major difficulty in fitness in soccer is the need to work with a large number of athletes and still take into account the principle of inter-individual variability. One way to mitigate this problem is to sub-divide the athletes into smaller groups according to certain characteristics and thus similar training needs. The cluster analysis (Cluster) K-means is a statistical tool that helps to organize data into structures with common features. Objective: To demonstrate the use of technology in K-means forming groups in a professional soccer team. Methodology: We used to analyze variables that translate the following aspects physical, aerobic power (VO2 max), aerobic capacity (anaerobic threshold), Agility (LITTLE, T & WILLIANS, A, 2005), body composition (% Fat and Lean Body Mass ), muscular power (vertical jump), speed test (30m). The sample was 35 professional athletes with a mean age 22:54 ± 2.9, n = 7 defenders, attackers n = 4, n = 2 goalies, n = 6 sides, socks n = 11, n = 5 flywheels. For analysis we used SPSS 17. Results: 3 groups were created using the F test for significant differences. Significant difference was observed for the groups created in Speed Threshold (G1: 13.80 ± 0.54, G2: 13.15 ± 0.47; G3: 12.65 ± 1.16, p = 0.011), Lean Body Mass (G1: 63.32 ± 2.88, G2: 72 ± 3.89 , G3: 61.92 ± 2.26, p <0.001), VO2 max (G1: 61.76 ± 2.62, G2: 59.64 ± 2.25, G3: 57.93 ± 3.9, p = 0.029), and vertical jump (G1: 41.4 ± 21.3, G2: 56.2 ± 5.89, G3: 51.03 ± 2.75, p <0.001), indicating that these groups should receive training for these different valences, as variables for Agility (G1: 5.46 ± 0.39, G2: 5.43 ± 0.37; G3: 5.39 ± 0:29 p = 0922) and% Fat (G1: 11.78 ± 1:41 G2: 12.78 ± 1.71, G3: 11.97 ± 1.88, P = 0317) groups are homogeneous and do not justify the creation of sub-groups. Conclusion: we conclude that the technique of K-means should be used in the formation of homogeneous groups. With this, showing a visual medium but not in the mathematical and scientific training organization.
Keywords: (K means cluster, Physical Training, Soccer)
Article: NOVACK, Luiz Fernando et al. Distribuição de subgrupos com base nas respostas fisiológicas em jogadores profissionais de futebol pela técnica K Means Cluster. Rev Bras Med Esporte [online]. 2013, vol.19, n.2, pp. 130-133. ISSN 1517-8692. http://dx.doi.org/10.1590/S1517-86922013000200012. http://www.scielo.br/pdf/rbme/v19n2/en_12.pdf.
E-mail: vitorbertolinascimento@yahoo.com.br
Introduction: A major difficulty in fitness in soccer is the need to work with a large number of athletes and still take into account the principle of inter-individual variability. One way to mitigate this problem is to sub-divide the athletes into smaller groups according to certain characteristics and thus similar training needs. The cluster analysis (Cluster) K-means is a statistical tool that helps to organize data into structures with common features. Objective: To demonstrate the use of technology in K-means forming groups in a professional soccer team. Methodology: We used to analyze variables that translate the following aspects physical, aerobic power (VO2 max), aerobic capacity (anaerobic threshold), Agility (LITTLE, T & WILLIANS, A, 2005), body composition (% Fat and Lean Body Mass ), muscular power (vertical jump), speed test (30m). The sample was 35 professional athletes with a mean age 22:54 ± 2.9, n = 7 defenders, attackers n = 4, n = 2 goalies, n = 6 sides, socks n = 11, n = 5 flywheels. For analysis we used SPSS 17. Results: 3 groups were created using the F test for significant differences. Significant difference was observed for the groups created in Speed Threshold (G1: 13.80 ± 0.54, G2: 13.15 ± 0.47; G3: 12.65 ± 1.16, p = 0.011), Lean Body Mass (G1: 63.32 ± 2.88, G2: 72 ± 3.89 , G3: 61.92 ± 2.26, p <0.001), VO2 max (G1: 61.76 ± 2.62, G2: 59.64 ± 2.25, G3: 57.93 ± 3.9, p = 0.029), and vertical jump (G1: 41.4 ± 21.3, G2: 56.2 ± 5.89, G3: 51.03 ± 2.75, p <0.001), indicating that these groups should receive training for these different valences, as variables for Agility (G1: 5.46 ± 0.39, G2: 5.43 ± 0.37; G3: 5.39 ± 0:29 p = 0922) and% Fat (G1: 11.78 ± 1:41 G2: 12.78 ± 1.71, G3: 11.97 ± 1.88, P = 0317) groups are homogeneous and do not justify the creation of sub-groups. Conclusion: we conclude that the technique of K-means should be used in the formation of homogeneous groups. With this, showing a visual medium but not in the mathematical and scientific training organization.
Keywords: (K means cluster, Physical Training, Soccer)
Article: NOVACK, Luiz Fernando et al. Distribuição de subgrupos com base nas respostas fisiológicas em jogadores profissionais de futebol pela técnica K Means Cluster. Rev Bras Med Esporte [online]. 2013, vol.19, n.2, pp. 130-133. ISSN 1517-8692. http://dx.doi.org/10.1590/S1517-86922013000200012. http://www.scielo.br/pdf/rbme/v19n2/en_12.pdf.
Comentários