Cognitive neural mechanism of sports competition pressure source

2.1 Cognitive neuroscience

The study of cognitive neuroscience aims to elucidation the brain mechanism of cognitive activities, that is, how the human brain USES its components at all levels, including molecules, cells, brain tissue areas and the whole brain to realize various cognitive activities[6]. Some branches of traditional neuroscience absorbed the theory of cognitive science and neuroscience, new technology, gradually formed a cognitive neuropsychology, cognitive psychology, physiology, cognitive psychology, cognitive neuroscience and computational neuroscience and cognitive neuroscience of each branch. In the late 1980s, the research of cognitive neuroscience has made remarkable progress in a short period of time, which has a great impact on the theoretical construction and research of traditional cognitive psychology and developmental psychology. The study of cognitive development is no exception[7,8]. Since cognitive developmental psychology and developmental neuroscience are interested in many common problems, the derived developmental cognitive neuroscience is attracting more and more attention and has become one of the hottest cross-disciplinary research fields.

Competitive stressors refer to “environmental needs directly related to competition”. How to systematically summarize and sort out these needs has become the primary task of competitive stressors research. According to the specific sources of environmental needs, it can be divided into organizational pressure source and competitive pressure source. 14 elite athletes from different sports organizations were studied, and the results showed that the subjects faced a lot of competitive pressure sources and organizational pressure sources. Using code analysis, the researchers will be competitive pressure source divided into seven aspects: preparation, such as inadequate physical preparation, injury (such as the opponent’s flagrant foul), pressure (such as the result of the game directly decides the personnel selection), competitors (if you don’t understand the opponent), self (e.g., in the game to show the perfect shape) and events (e.g., performing complex technical movements), superstition, such as not take “lucky” equipment to match). And organizational pressure source is classified as four aspects: environmental issues, such as personnel selection unfair, the lack of financial support, training, facilities), personal problems, such as to his own expectations too high, the lack of goals and direction, and state for their high hopes), leading questions, such as tension coaches and athletes, coaches are not good at communication, coach too bossy, the coach’s professional skill not good), team issues (such as the lack of communication between teammates, his teammates adventurous, management did not fulfil his obligations, its role is not clear). The study also found that athletes reported more accurate amounts of organizational stressors and more diverse forms of performance than competitive stressors.

2.2 K-means clustering

Clustering is unsupervised learning that groups similar objects into a cluster. The clustering method can be applied to almost all objects. The more similar the objects in the cluster are, the better the clustering effect will be. K in the k-means algorithm represents the clustering of K clusters. Means to take the mean of the data values in each cluster as the centre of the cluster, or the centre of mass. That is to say, the centre of mass of each class is used to describe the cluster. The biggest difference between clustering and classification is that the goal of classification is known in advance, while clustering is different. Clustering does not know what the target variable is in advance, and categories are not defined in advance like classification. Therefore, clustering is sometimes called unsupervised learning. Cluster analysis tries to group similar objects into the same cluster and classify dissimilar objects into different clusters. Therefore, it is obviously necessary to find an appropriate similarity calculation method. There are many known similarity calculation methods, such as Euclidean distance, cosine distance, hamming distance and so on. In fact, we should select the appropriate similarity calculation method according to the specific application.

It is widely used as the k-means algorithm and is sometimes referred to as the Lloyd algorithm (especially in computer science) [9,10]. Given the initial k mean points m1(1),…,mk(1)false, the algorithm follows the following two steps:

I) Assignment:

Each observation was assigned to the cluster to minimize the sum of squares and (WCSS) in the group.

Each xp false is assigned to a defined cluster St false, although in theory it may be allocated to two or more clusters.

II) Update:

For each cluster obtained in the previous step, the center of the observed value in the clustering is used as the new mean point.

Because the arithmetic average is the least square estimation, this step also reduces the value of the sum of squares and (WCSS) in the target function group. This algorithm converges when the distribution of observations is no longer changing. Since the two steps of alternating will reduce the value of the target function WCSS, and the allocation scheme is only limited, the algorithm is bound to converge to the optimal solution of some (local).

3.1 Data process

It consists of 23 assessment indicators, used for the frequency, intensity and duration of the movement of athletes asked a series of factors such as profiling. The questionnaire is translated from this questionnaire. By random questionnaire survey form, due to the cooperation with the school sports institute, the data source provided by the school sports institute, and the original data source is to the Wuhan sports college sports teams from 500 different levels of different hierarchical levels of the athletes have done a complete set of sports of psychological evaluation questionnaire data, 500 athletes of different projects, different grades and different ages filled in the questionnaire. A total of eight questionnaires, 22 kinds of attributes, and the questionnaire is divided into five grades, the psychological characteristics of general use: do not conform to the part, not completely conform to, in line with, most accord with, fully comply with the rank; General use of behavioral action: never, occasionally, sometimes, often, and always, the two are not fundamentally different. After obtaining the data, the questionnaire is quantified using 1 to 5 points.

1. Stress source questionnaire (# 11, 12, 16, 18, 20, 23); Logistics and Operations, (# 3, 5 6, 7, 14, 17, 19, 21, 22); Team of athletes (# 1, 8, 10, 13), mainly from team teammates, the relationship between team as the atmosphere of whole on its positive or negative factors; They are shown in table 1:

1. Mental toughness questionnaire.

Confidence dimension: (# 13, 5, 11, 6, 14, l); Firm dimension, (# 3, 12, 8, 10); Control dimensions, (# 2, 4, 9, 7). They are shown in table 2:

1. Feelings questionnaire

Confidence dimensions (#1, 5, 9, and 13); Vitality dimension, (#2, 6, 10, 14); Dedication dimension, (#3, 7, 11, 15); The enthusiasm dimension, (#4, 8, 12, 16). They are shown in table 3:

3.2 K-Means process

1. Data for clustering

In chapter 3, the questionnaire is divided into: competition pressure source text, psychological resilience questionnaire, and exercise feeling questionnaire. Among them, the questionnaire is the most important in the competition pressure source. In addition, each questionnaire is divided into several sub-dimensions, and the detailed questionnaire and the subdivision dimensions are shown in table 4:

After the evaluation of nearly 500 athletes, the comprehensive data are analyzed and the relevant clustering information is obtained. Due to the original questionnaire evaluation ranks generally divided into five segments (fully meet, compared with, general, more do not conform to, completely is not in conformity with the), so, get data raw score, score for each question 5 points, cumulative scores for the current breakdown dimension scores.

According to the improved algorithm, after the data is cleaned and transformed, the clustering analysis should be carried out. Firstly, the contour coefficient should be used to calculate the approximate number of clusters K, figure 4. 3 represent the value curve of the contour coefficient (re) between 2 and 100. By figure 4 _3, the clustering number when 2 profile coefficient, maximum value indicating the clustering number when 2 effect is good, but this is not the final result, we can start with hierarchical clustering algorithm for clustering, and based on the improved algorithm, and calculate the next K-means the initial clustering center, convenient for subsequent calculations [11,12].

After calculation, the initial clustering center can be obtained as shown in table 5:

1. Result analysis

Improved hierarchical K - Means algorithm to cluster the data sets, the source of competition pressure, social support, the athlete burnout 22 dimensions such as clustering, get the final clustering center as shown in table 5.

In figure 1 and figure 2 it can be seen that the first class occupies the most (blue line);

Figure1

Silhouette Coefficient

Figure 2

The pressure distribution of layered algorithm in sports competition

Category 2 second (green line), the third class at least (red line) type 1 and type 2 is bigger, the overall similarity part attribute is different, and the third class accounted for less than the total number, but have obvious difference with the class 1, 2, means that this part of the athletes psychological level is poorer (overall score is low, the pressure source does not conform to, less pressure).