Impact Analysis of Basketball Exercise Strength Based on Machine Learning in the Mental Health of College Students.

In the current environment of globalization, the communication between people is gradually getting closer, and the society is becoming more and more complex. With the continuous development and progress of science and technology, people are more skilled in applying science and technology to their own concerns. College students are about to enter the society, will feel multiple pressure from family, school, and society, study and life problems will gradually convert into mental health problems, and we need to use machine learning basketball exercise to positively affect the mental health quality of college students. The improvement of living conditions makes people pay more attention to their physical and mental health, and learn to use machine learning sports reasonably, not only basketball exercise, to improve mental health diseases. However, we need to use machine learning to identify the different effects of different basketball exercise intensity on mental health, in order to ensure that the most appropriate basketball exercise intensity brings good aspects to the mental health of college students. Through the investigation and data sampling, it can be concluded that the machine learning-based basketball exercise intensity has a positive impact on the mental health of college students.

1. Introduction

With the continuous development and progress of society, people pay more attention to their physical and mental health, not only due to physical health but also pay more attention to mental health. In the social environment with the rapid development of computer networks, the use of machine learning for emotion recognition has attracted wide attention. People care about the body of their family and friends. Emotional recognition has very broad application prospects in many fields of medical care, online education, and other fields. The designed emotion recognition algorithm is essentially a multi-classifier integration algorithm to achieve better results by integrating classical machine learning classifiers. After the data information collected by the sensor is processed centrally, it is inserted into the algorithm of emotion recognition for calculation. The calculated results are analyzed and the results are displayed, and the long and short results are summarized to provide guidance for the subsequent research path and direction [1]. This article introduces the data types and sources of machine learning for us, introduces a variety of machine learning algorithms, combined with the actual situation in the field of mental health, improve different machine learning algorithms, recognize its advantages and disadvantages, combined with the advantages of traditional psychological tools and machine learning to machine learning in the field of mental health. Finally, it is summarized to play a positive role in effectively solving the prevention of psychological problems [2].

Basketball as a kind of aerobic exercise is an important way to improve mental health problems and maintain mental health problems. After years of research, many research results have shown that basketball can improve psychological problems. This article is to study whether the intensity of basketball exercise has an impact on the mental health of college students. The machine learning method and statistical research results show that the different intensities of basketball exercise will indeed have different positive effects on the mental health of college students. But we need more experiments to support this result [3]. In the moderate exercise intensity group and no exercise group, they had interpersonal social interaction, anxiety sensitivity, more than depression paranoia, and psychosis, which was in line with the comparable conditions in statistics. Through the experiment, small-intensity basketball has improved interpersonal social fear and other symptoms to a certain extent [4]. It discusses the influence of basketball exercise intensity on the mental health of college students, provides guidance methods and related research materials for college physical education activities, and concludes that basketball sports of different intensity can improve the mental health of college students [5].

Sports and mental health are interrelated. They restrain and influence each other. Strengthening sports basketball can enhance students' interest and learning motivation, and improve the mental health environment of college students. Effective exercise can make college students effectively control their psychological emotions and pressure, and also improve their interpersonal relationships and adaptability to the environment. Therefore, sports have a certain impact on the mental health of college students [6]. The article thinks that sports have a positive guide to college students' psychological problems. This phenomenon will produce different reactions because of the different intensity, frequency, and even time of exercise. Later, the calculation method of machine learning should be used to study the influence mechanism of physical activities on college students [7].

This study discusses the influence of low- and medium-intensity basketball on self-esteem and mental health and provides the basis for physical education and mental health education. There was no difference in self-esteem before and after intervention in the control group. Low-intensity basketball activities have an impact on the mental health of students in the experimental group. After the intervention, students in the moderate intensity group scored lower in obsessive-compulsive disorder, interpersonal sensitivity, depression, phobia and anxiety, mental quality, and general scores. In conclusion, middle- and low-intensity basketball is beneficial to boys' self-esteem and mental health, and the middle-intensity group is better than the low-intensity group [8].

After years of intensive study, the most pressing problem that college students must face is employment. With the sustained development of China's economy and the accelerated pace of social progress, the society's demand for talents comprehensiveness and adaptability is increasing. However, under the influence of long-term exam-oriented education, college students are only satisfied with the required courses at present, resulting in a lack of professional knowledge and the ability to solve practical problems, narrow-mindedness, poor flexibility, creativity, and operational ability. With the significant difference between campus and society, most college students will feel pressure from all aspects when they first enter the society. For example, they lack effective learning methods, have strong social responsibility and self-study ability, have poor understanding and mastery of professional knowledge, poor sense of unity, poor social practice and operation ability, and poor cognitive ability. Moreover, they often have great communication barriers and cannot correctly evaluate themselves and others [9].

Strengthening sports can improve students' physical quality. Physical and mental exercise is to achieve physical health and mental happiness through physical activities. Academics have long studied the relationship between physical exercise and mental health and reached a consensus that physical exercise has a positive promoting effect. College students are the future talents of this country. Under the national policy of attaching importance to talent building, more and more attention is paid to students' mental health [3]. This study compares the students who use mastery learning methods with those who use nonmastery learning methods [10].

This study aimed to explore the influence of different sports events, intensity, and time on college students' anxiety and depression. College students' sports participation, perceived motor score, and self-rating anxiety scale were discussed in depth. The relationship between the intensity and frequency of basketball exercise and mental health anxiety and depression was studied. Anxiety and depression exist among most ordinary students. College students can take different sports to improve their mental health [11]. Basketball can effectively improve the emotional state and physical level of college students and promote the mental health of male college students [12]. College students' mental health education is playing an increasingly important role in higher education. How to cultivate good mental health through physical education is a difficult problem faced by all educators in China. Let students know more about basketball, and by participating in the basketball team experience, they can improve their social skills, so that college students can better deal with all kinds of social relationships in campus life and study, and build a good mental health quality [13]. Traditional exercise prescription is for physical health, but many current college students need mental health. We should pay attention to how to improve the traditional sports prescription and strengthen their mental health adaptability. Take basketball as an exercise form of mental health prescription [14].

2. Machine Learning Algorithm

2.1. KNN Algorithm

KNN algorithm is one of the classical algorithms in machine learning [15]. Its core idea is that if the K most adjacent samples in the feature space belong to a certain type, the sample also belongs to the current type and has the characteristics of that type. The formula for calculating the distance between the two samples iswhere P is the variable parameter, when P = 1, the formula represents the Manhattan distance; when P = 2, the formula represents the European distance; when P⟶ the formula represents the Chebyshev distance.

2.2. Random Forest Algorithm

Random forest algorithm cannot be separated from the decision tree. Random forest algorithm consists of multiple decision trees, and there is no connection between different decision trees. The prediction results of the random forest algorithm are the results of the classification displayed by most decision trees. The random forest algorithm is more accurate and stable compared to the results predicted by a single decision tree.

2.2.1. Detection Process of a Random Forest

The detection of random forest is using a single decision tree to detect characteristic samples step by step, and then uses the minority-obeying majority voting mechanism to judge the final boundary result. The binary segmentation function is defined as follows:

In the formula h(x, θ)—The binary segmentation function at the current node;

θ —Characteristic properties and the corresponding threshold of the sample x stored at the node j;

τ—Threshold value of property k for sample x at node j.

2.2.2. Training Process in the Random Forest

Each decision tree was trained independently. The training sample set of the decision tree includes feature samples and the corresponding artificial labeled samples. Use the Gini exponentially split feature attributes to generate decision trees. The samples are processed as follows:

In the formula S—The training sample S ∈ X × Y, X at the node j represents the feature sample, and Y represents the marker sample;

Gini—The Gini index at the node j;

S —A nonboundary sample at the current node j;

S —Samples belonging to the motion target boundary;

Redefining the Gini exponential splitting function for the dichotomy case:

In the formula H(S)—Split indicators at the nodes j calculated using the Gini index;

y—is a certain structured label, which has a variety of structures;

p —Probabilities corresponding to the structured label 1;

Split—Formula representation of the second half of the information gain standard function.

2.3. ID3 Algorithm

The ID3 algorithm is the earliest decision tree algorithm. The information gain is calculated by the following formula:where X is the set of all training samples located at the current node, n is the n species of classification corresponding to the sample set, p is the i corresponding probability, m is the m value cases for the attribute A, and p is probability that the attribute A is worth j.

2.4. C4.5 Algorithm

ID3 algorithms are prone to fitting with few training samples and more algorithm parameters. In view of this deficiency, Quinlan has improved the ID3 algorithm, and the improved algorithm is called the C4.5 algorithm.

The C4.5 algorithm can disperse the continuous attributes based on the ID3 algorithm and can handle the training data with missing attribute values. C4.5. Definition of the algorithm:where m is the m values of the attribute A, Split Info(X) is the amount of information that A contains in the current sample X, and Gain Ratio(A) is information gain ratio corresponding to attribute A.

2.5. CART Algorithm

The CART algorithm adopts the Gini exponential splitting properties, and the splitting criteria are as follows:

In the formula Gini(X)—The Gini index value of the current sample set X;

p —The probability corresponding to the sample set category i;

m—Attribute A has m values. The current sample can be divided into m subsets;

N —Total number of samples corresponding to the kth subset;

Gini(X)—Gini split values for split by attribute A.

2.6. Expectation-Maximization Algorithm

The EM algorithm is an effective method to find maximum likelihood estimation or maximum posterior estimation with hidden variable models [16]. It is widely used in natural language processing, psychology, quantitative genetics, and other problems. The EM algorithm was proposed in 1977 by Arthur Dempster, Nan Laird, and Donald Rubin. The method is simple and effective in operation [17].

Log-likelihood function for a given Gaussian mixture distribution is as follows:

Seek the maximum likelihood estimate of the parameter θ, namely:

Using the EM algorithm to estimate the parameters simplifies the problem.

First, finding the partial guide for the mean μ of each component of ln  p(X|π, μ, ∑) relative to the mixed Gaussian distribution can obtain:

Multiplies the above equation by ∑−1 and rearranges it to obtainwhere N is defined as

Through the calculation of the above formula, the mean μ of the Gaussian component is the case of the Gaussian mixture distribution. Then, following the same method, ln  p(X|π, μ, ∑) for each μ, you can obtain

Observing the equation above shows that the form of the solution of the parameter μ in a mixed Gaussian distribution is similar to the solution of the covariance in a single Gaussian distribution.

Finally, the corresponding per-mixing coefficient π is obtained by maximizing ln  p(X|π, μ, ∑). Specifically, the constraint of π can be incorporated into the objective function by introducing the Lagrangian coefficient, subsequently maximizing the new objective function:

Similarly, it can be obtained by finding the partial guidance for π:

Subsequently, using the characteristic of ∑π=1 to sum the probability π of each component, you can simplify the following results: λ=−N. Bring λ=−N into the original to eliminate, and rearrange, and finally obtainwhere π is the mixing coefficient for the kth component in the mixed Gaussian distribution.

3. Status Quo of Basketball Sports in China

As is known to all, basketball has become a well-known popular sport in China. People from primary school students to middle-aged and elderly people participate in basketball. This sport spans all ages and gender races, and everyone gets together to play basketball.

With the development of artificial intelligence and the progress of machine learning methods, in recent years, domestic basketball research has gradually developed in the direction of combining theory and practice. Compared with foreign anaerobic basketball training, aerobic training is more inclined in China. In recent years, with the help of machine learning science concepts and equipment, China has enriched basketball training methods. However, the index system for assessing the level of basketball mainly lies in special speed, strength, and endurance. The research on basketball training in China mainly focuses on physical quality, thus ignoring the mental health state, but the real basketball sport should lie in the comprehensive combination of form, function, and quality, and then the physical fitness level combined with people's health. The difference in basketball exercise intensity not only brings us physical changes but also is a process of physical and mental pleasure, which can shed negative energy and negative emotions.

In the physical education activities of colleges and universities, basketball has always been a hot course, which shows the love of college students for basketball, and observing every college basketball court, is always people, full of, filled with the power of youth. Under the multiple pressure of contemporary college students, basketball is no good way to relieve pressure and relax, so basketball exercise under machine learning can really improve the mental health quality of college students.

Guo Haoran put forward the research [18] on the teaching and education of basketball courses in colleges and universities in the Experimental Research of Introducing Physical Fitness Education on Improving the Teaching Effect of Basketball Public Courses in Colleges and Universities. It means hoping to combine basketball with physical education practice, Add basketball education to physical education, While improving the quality of physical education in colleges and universities while promoting the benefits of basketball exercise, Has the following benefits: (1) is a major breakthrough to improve the traditional sports teaching, Better improve the traditional basketball teaching; (2) Improving the focus of traditional physical education teaching, Pay more attention to the physical and mental health of college students; (3) Better popularization of basketball exercise brings not only the physical benefits, It can also promote the development of mental health.

4. Mental Health Analysis

4.1. Implications of Health and Mental Health

The United Nations Health Organization defines health as: “Health is not only without physical defects, but also with a complete physical, psychological state and social adaptability.” Thus, a healthy person is not only a healthy body but also needs a healthy mind. Although there are many controversies in society, even if each person's ideas are different, but mental health is a part of human health.

So what to measure the mental health of college students. Combined with the relevant research at home and abroad, and starting from the actual conditions of the psychological development and mental health of Chinese college students, it should be reflected from the following aspects: (1) healthy and stable mood; (2) strong and brave quality; (3) positive social adaptability; and (4) personality integrity and behavior coordination [19].

4.2. Physical Education Teaching and the Mental Health of College Students

Sports teaching relationship with college students' mental health [20], the characteristics of college students' psychological development, state, motivation development level and master sports knowledge, sports skills, and the main basis of design, arrangement of physical education teaching, reasonable arrangement, and scientific organization sports teaching can make some possibility of psychological development become reality to promote the healthy development of students' psychology. With the progress of sports psychology and the popularization of mental health knowledge, people also gradually realize that physical exercise can not only strengthen physical fitness but also promote mental health.

Physical education teaching is to consciously adjust students' emotions in the process of participation, enhance the communication between students, enhance students' confidence, cultivate the courage to struggle, cultivate a strong will quality, as well as physical education teaching to make students get healthy ideas, and establish healthy behavior. Its main aspects are to promote the development of students' cognition, promote the development of students' emotions, promote the development of students' will and quality, promote the development of students' personality, slow down stress and improve mental health [21].

4.3. Current Quo of Mental Health of College Students

In the current rapidly developing information age, college students have to bear the triple pressure from their families, schools, and society. According to relevant statistics, about 20% of college students have different degrees of mental health problems, which we should take corresponding positive measures so that college education can cultivate high-quality talents to meet the social development and needs.

However, because of people's high expectations and requirements for college students, their self-concern and life goals, the mental health problems of college students are significantly higher than other groups of the same age, so they must be paid great attention to Colleges and universities are related to the mental health problems of college students. Schools should proceed from reality, carry out a large-scale psychological investigation and in-depth and meticulous research and analysis, use scientific sampling and standards to judge the mental health status of college students, and provide accurate and effective teaching methods for the development of mental health education [22].

4.4. Impact and Effect of Sports on the Mental Health of College Students

As we all know, physical exercise can effectively improve the sensitivity and coordination of the human body. But in fact, this is only one aspect of physical exercise in colleges and universities. In physical exercise, students get exercise not only physical exercise but also mental health exercise, like physical exercise, which is a process of self-improvement in continuous understanding and self-improvement. College students can improve their thoughts, morality, will, emotion, and other aspects through physical exercise. Basketball is an important branch in physical sports [23].

5. Example Analysis

5.1. Experiment Preparation

The 40 students from Chongqing University were selected. Among them, 20 basketball (10 male and female) and 20 ordinary college students (10 male and female) were selected for physical testing and psychological testing.

First, understand the acceptance intensity of basketball at each level and then make basketball prescription for ordinary college students; the basketball teacher is responsible for the guidance, three times a week, 1.5 hours, for 8 consecutive weeks according to the set intensity of fitness activities.

The following data were obtained from the identification survey of the value of physical exercise among 200 random students (Table 1) and the degree of daily physical and mental feelings of college students (Table 2).

Table 1

College students' identification of the value of sports (unit: %).

Content	In full agreement	Largely agree	Disagree	Pussyfoot	Total
Sports can promote physical and mental health	76.8	22.8	0.4	0	100
Sports can enrich your spare time life	65.0	33.3	1.2	0.4	100
Sports can develop good living habits	46.3	46.7	3.7	3.3	100
Sports can relax	44.7	54.9	0.4	0	100
Sports can cause fatigue and affect learning	2.4	8.5	80.9	8.1	100

Table 2

Daily physical and mental feeling degree of college students (unit: %).

	Insentience	Sometimes feel	Often feel	Total
Physical fatigue	8.5	81.7	9.3	100
Mental fatigue	10.6	73.2	16.3	100
Feel underexercised	20.3	68.3	11.4	100
Feel that obesity is happening	44.7	37.4	17.9	100

It can be seen from Table 1 that most college students have a positive attitude towards physical exercise, and the positive influence of college students on physical activities is relatively high, indicating that college students have a relative sense of identity for the improvement of physical and mental health of physical exercise, including basketball exercise.

It can be seen from Table 2 that college students usually feel physical fatigue and mental fatigue in daily life. In addition, most students think that physical exercise can bring changes, which proves that contemporary college students have independent personality and positive attitude towards life.

5.2. Basketball Quality Test for Students

As for the statistics of the basketball level of the selected students (Figure 1), it can be seen that very few college students participate in the basketball level and obtain the basketball level certificate.

Figure 1

Basketball level.

Considering the difference between basketball special students and ordinary college students at the speed of 5 km/h, Figure 2 shows the right step length of basketball special students is not significantly different from that of ordinary college students, but the left step length of ordinary college students is slightly larger than that of special basketball students. It can be seen that basketball exercise for step length and step width change is not big.

Figure 2

Step length and width of each group at 5 km/h (m).

Figure 3 is a comparison of the gait cycle of basketball students and ordinary college students at 5 km/h. The gait cycle of basketball students is less than that of ordinary college students. After a long time of basketball training, basketball exercise has a certain effect on the gait cycle.

Figure 3

Gait cycle of each group at 5 km/h (seconds).

Figure 4 shows the statistics of step frequency speed change for special basketball special students and ordinary college students. Step frequency refers to the number of steps per unit time, expressed as in (times/per minute). At 9–13 km/h, with the increasing speed of basketball special students and ordinary college students, the pace frequency also showed an increasing trend. However, the pace frequency of basketball special students is more stable, which shows that basketball exercise has improved the stability of human sports.

Figure 4

Changes in step frequency with speed in each group (times/minute).

5.3. The Impact of Basketball on College Students' Body

5.3.1. Effect of Basketball on Heart and Lung Function

Table 3 is an 8-week basketball training for ordinary college students. Through basketball training, the function of the body has a significant impact. Before and after 8 weeks of basketball training, the indicators of cardiopulmonary function of college students after exercise are significantly improved compared with those before exercise. Basketball has a positive impact on improving the body's heart function and promoting the improvement of vascular function.

Table 3

Comparison of heart and lung function before and after basketball exercise (20 people).

Index	Front	Behind	P
Pulse (b/min)	80.84 ± 2.72	73.94 ± 2.10	<0.05
Diastolic pressure (kPa)	73.74 ± 2.29	67.05 ± 1.65	<0.05
Systolic pressure (kPa)	121.53 ± 2.68	116.89 ± 1.89	<0.05
Step index	57.17 ± 1.44	59.30 ± 1.37	<0.05
Vital capacity (ml)	4480.79 ± 134.65	4565.78 ± 127.96	<0.05
Spiro-index (ml/kg)	68.15 ± 2.48	69.40 ± 2.32	<0.05

5.3.2. Influence of Basketball Exercise on the Physical Quality of College Students

As can be seen from Table 4, all the quality tests after exercise are higher than those before exercise. After exercise, the speed of running and 50 meters is significantly higher than before exercise, but the speed of 1000 meters is not much different, indicating that the sensitivity and speed of college students need to be improved, but the endurance is difficult to change. The improvement of pull-up quality shows that the upper limb strength quality of college students needs to be improved. Eight weeks of basketball exercise has a certain impact on the speed, endurance, sensitivity, and the explosive power of the upper and lower limbs of college students. The basic physical quality is improved greatly, and the effect is remarkable.

Table 4

Comparison of various qualities before and after basketball exercise (20 people).

Index	Front	Behind	P
Shuttle run (s)	11.11 ± 0.16	10.54 ± 0.11	<0.05
Push-up (individual)	27.74 ± 2.60	30.21 ± 2.24	>0.05
Pull-up (individual)	3.74 ± 0.61	6.68 ± 1.34	<0.05
Grip (N)	50.39 ± 1.22	54.85 ± 1.29	<0.05
Bending (cm)	10.14 ± 1.07	10.21 ± 1.70	>0.05
Vertical jump (cm)	50.37 ± 1.98	56.10 ± 1.57	<0.05
Long jump (cm)	2.51 ± 0.04	2.61 ± 0.04	<0.05
50 meters (s)	7.50 ± 0.11	7.30 ± 0.11	<0.05
1000 meters (s)	3.97 ± 0.18	3.73 ± 0.17	>0.05

5.3.3. Impact of Basketball Exercise on the Physical Shape of College Students

As can be seen from Table 5, before and after the basketball exercise, the muscle and bone weight of college students increased, the muscle volume increased, and the bone density and bone mass increased. Body mass index BMI: 18.5 < BMI < 23 was within the normal range.

Table 5

Comparison of each circumference, skinfold thickness, and derived index before and after basketball exercise (20 people).

Index	Front	Behind	P
Stature (cm)	174.46 ± 1.37	175.26 ± 1.34	<0.05
Weight (kg)	67.13 ± 3.09	66.91 ± 2.90	>0.05
Bust (cm)	91.52 ± 1.85	89.84 ± 1.75	<0.05
Upper-arm circumference (cm)	30.60 ± 2.55	27.43 ± 0.64	>0.05
Waistline (cm)	76.94 ± 2.59	74.78 ± 2.10	<0.05
Hipline (cm)	93.49 ± 1.55	92.96 ± 1.46	>0.05
Thigh circumference (cm)	53.57 ± 1.35	53.48 ± 1.33	>0.05
Three humerus head (cm)	10.15 ± 1.34	9.27 ± 1.14	<0.05
Loin (cm)	13.30 ± 2.01	12.24 ± 1.88	<0.05
Belly (cm)	10.58 ± 1.40	9.66 ± 1.26	<0.05
BMI (kg/m2)	21.97 ± 0.84	21.71 ± 0.79	<0.05
WHR	0.82 ± 0.016	0.80 ± 0.012	<0.05

5.4. The Influence of Basketball Exercise on Psychological Problems

The results of Table 6 were obtained through the investigation:

Table 6

Comparison of student SCL-90 test results before and after basketball exercise.

	Before the experiment	After the experiment	P
Somatization	1.517 ± 0.407	1.489 ± 0.334	>0.05
Forced symptoms	1.719 ± 0.452	1.576 ± 0.487	<0.05
Interpersonal relation	1.699 ± 0.324	1.424 ± 0.289	<0.05
Depressed	1.531 ± 0.419	1.389 ± 0.392	<0.05
Anxious	1.395 ± 0.281	1.282 ± 0.269	<0.05
Hostile	1.282 ± 0.195	1.165 ± 0.264	<0.05
Terrifying	1.484 ± 0.262	1.416 ± 0.261	>0.05
Bigoted	1.197 ± 0.287	1.101 ± 0.236	<0.05
Psychiatric sex	1.208 ± 0.218	1,132 ± 0.212	<0.05
Add	1.412 ± 0.365	1.394 ± 0.313	>0.05

Through the analysis of the results in Table 6 and Figure 5 above, we can see that the basketball exercise in machine learning has a positive impact on the mental health problems of college students, especially the differences in interpersonal relationships. It shows that the exercise of different intensities of basketball has a significant effect on improving the mental health level of college students. Through the basketball exercise activities, the physical quality and the mental health of college students have been improved. Using machine learning methods to achieve different degrees of basketball training intensity can indeed improve the mental health quality of college students [24].

Figure 5

Comparison of SCL-90 test results of students before and after exercise.

6. Conclusion

In view of the impact of basketball exercise on the mental health of college students, the introduction of machine learning is preliminarily studied, but further research is still needed. The application of machine learning technology methods, especially decision tree methods, has just started in the field of mental health research and faces many uncertainties and challenges, but it has broad application prospects.

Aiming at the forefront of machine learning development, we constantly introduce the latest machine learning technology methods. Taking the field of mental health research as the focus of research, we can compare and analyze the characteristics of different basketball exercise intensity through simulation experiments, so as to further improve the ability of machine learning to judge and simulate the strength of basketball exercise.

In the face of the doubts in the field of traditional basketball education, to study the interpretability of deep learning. Based on domain knowledge, interpretable machine learning models should be established to promote the development of interdisciplinary research in the field of basketball sports education and mental health research.

For the basketball exercise fully reasonable use of machine learning technology and algorithm and exercise intensity of test and experiment, for the changes in mental health analysis, take the optimal scheme, not only can improve the students 'sports and learning enthusiasm, but can also bring positive influence for college students' mental health.