Research On Revising A Scoring Table Based On Neural Network Model

The decathlon includes 100m, long jump, shot put, high jump, 400m ,110mH, discus throw, pole vault, javelin throw, 1500m,and is challenged to combine excellent physical power, explosiveness, technical(psychomotor) skills and endurance. The idea of the combined event goes back to the ancient Greeks, who introduced the pentathlon-running, discus and javelin throws, jumping and wrestling. The winner was considered the most complete athlete and was accorded almost godlike status.As the core idea of the decathlon is all-roundness, a well-founded scoring model is necessary to combine the performances in the various disciplines into a total score. For this purpose official scoring tables have been developed for each of the events. Over the years, these tables have been the subject of extended debates about their fairness and validity. On several occasions amendments have been made in order to remove manifest imbalances between disciplines. These may easily arise because of new training approaches or new materials. Indeed, some disciplines steadily evolve while others tend to stagnate.As will be show in this paper, the current scoring method appears to be cursed with unacceptable bias and needs a conceptual revision. In this paper, extensive reading is on the basis of literature. Literature data, curve fitting, progressive scoring and neural network modeling method is used to revise the scoring table in order to make it more reasonable, so that the decathlon sport can be toward the direction of comprehensive development.As is followed, this paper gets some conclusions based on some study:1. In the decathlon scoring tables, there are three main problems: the increase of point fails to reflect the principle of progressive scoring; fairly bigger difference existed between the lowest and the highest points; the degree of difficulty of scoring for different events was inconsistent; for some events, the scoring table was insufficient.2. Progressive scoring is relatively reasonable to revise a scoring table. By some experiments, the tendency between the sports score and score point increases quickly, which can reflect the fact that the difficulty will increase drastically when decathlon athletes want to increase the score point.3. BP neural network model is a more rational form to revise the scoring table, which adjusts the weight right with gradient descent, but convergence rate is slower.