FYP Proposal Slides

INJURY PREDICTION OF NBA PLAYERS WITH REGRESSION NEURAL NETWORK.

[Audio] The study presents an injury prediction model for N-B-A players using regression neural network. The research objectives are to understand the factors contributing to injuries in N-B-A players and to develop a model that can predict the likelihood of injury. The presentation will include an introduction problem statement literature review methodology and references..

[Audio] Our research on injury prediction of N-B-A players using regression neural network has been built on the latest advancements in machine learning opening up new possibilities for injury prevention research. Injuries are a common occurrence in the N-B-A and can have a significant impact on team performance and player careers. Preventing injuries is crucial for achieving peak performance in sports and maintaining good health. We are excited to share our findings with you..

[Audio] We create an injury prediction model for N-B-A players using S-V-M and R-N-N--. Our extensive dataset includes player information and performance metrics enabling us to accurately forecast potential injuries. We inform team physicians and trainers about injury risks and provide valuable insights to improve player safety and performance..

[Audio] The basketball industry is facing several challenges one of which is the high frequency of season-ending injuries which can significantly impact player performance and team success. Injury predictive models are lacking making it difficult for teams to identify potential injuries and take preventative measures. To address these problems a regression neural network can be used to improve injury prediction accuracy and reduce the impact of injuries on player performance and team success. This approach can be applied in the N-B-A to improve the accuracy of injury prediction and ultimately aid in the prevention of injuries leading to better player performance and team success..

[Audio] Our injury prediction study for N-B-A players has resulted in a precise injury predictive model the identification of significant risk factors and the evaluation of the models' performance. We have developed a precise injury predictive model through our regression neural network. We have identified several significant risk factors that can lead to injuries in N-B-A players. Our evaluation of the models' performance has shown that our regression neural network outperformed the other models. Our study has provided valuable insights into injury prevention for N-B-A players and we hope that our findings can help improve the safety and well-being of athletes at all levels..

[Audio] Our literature review focuses on machine learning techniques used in sports injury prediction specifically Regression Neural Networks (R-N-N--) for injury prediction. Integrating multiple data sources as emphasized by Jordan & Mitchell (2015) is crucial for increasing prediction accuracy. ruddy and others (2018) highlights that no single technique can consistently outperform others. Regression Neural Networks (R-N-N--) as proposed by Goodfellow and others (2016) have shown to be useful for injury prediction as they can learn complex patterns and relationships. Our research specifically focuses on lower extremity muscle strain injury prediction showcasing the potential of neural networks in building an accurate prediction model..

[Audio] XGBoost is a highly effective technique for building a strong model by iteratively adding weak learners. It is an ideal choice for sports injury prediction because of its ability to handle large datasets and feature interaction. Rommers and others (2020) and Kuhn & Johnson (2013) have highlighted the effectiveness of XGBoost in accessing injury risks in elite-level youth football and handling missing values and feature selection respectively..

[Audio] We present a methodology for injury prediction of N-B-A players using a regression neural network. Our approach emphasizes data collection variable selection and outcome measurement. We preprocess the data to improve accuracy and interpretability of the model. We then implement the model and select the most suitable one..

[Audio] Our data is from kaggle.com covers the period from 2002 to 2017 and includes injury dataset demographic characteristics and performance metrics. Our variables and outcome include season-ending injury age position played weight height free throw rates three-point rates points rebounds and more. We have preprocessed our data by filtering relevant data between 2002 and 2017 merging the injury dataset with players’ season stats and normalizing continuous variables..

[Audio] Regression neural networks can predict injuries in basketball players by utilizing an input layer hidden layers and an output layer. The networks are trained using backpropagation and gradient descent algorithms with mean squared error as the loss function. This technique enables the achievement of non-linear relationships between variables making it an effective tool for injury prediction in basketball players..

[Audio] We can improve classification accuracy and reduce errors by using support vector machine (S-V-M--) with radial basis function kernel. This technique can handle non-linear relationship between data which is important for accurately separating data into distinct classes. By using S-V-M with radial basis function kernel we can identify the best hyperplane for separating data into distinct classes resulting in improved classification accuracy and reduced errors..

[Audio] XGBoost a popular machine learning algorithm is utilized for injury prediction for N-B-A players. The utilization of weak learners in XGBoost creates a strong predictive model. Through the use of regression neural network we can improve the performance of our injury prediction model by making more accurate predictions..

[Audio] Discuss injury prediction of N-B-A players using a regression neural network. This method ensures a balance between recall and precision. The effectiveness of this approach has been shown in previous studies..

[Audio] Our focus is on injury prediction of N-B-A players using a regression neural network. We have validated our model on different time periods to ensure its robustness. We have compared our model results to detect overfitting and underfitting. Model interpretability is also a priority for us and we have used S-H-A-P and feature importance to identify the most influential factors in our model. S-H-A-P value can help us understand the impact of each factor on injury prediction..

[Audio] To ensure the accuracy of injury prediction for N-B-A players We will check the dataset and apply each of the three machine learning techniques to it. We will evaluate and compare the performance of each model to determine which one is the most effective for injury prediction..

REFERENCES. UNIVERSITI PUTRA MALAYSIA. 15.

[Audio] In this presentation we will discuss the limitations of traditional injury prediction methods and introduce a Regression Neural Network. Our goal is to demonstrate how this technique can provide more accurate and reliable predictions leading to better injury prevention and management strategies. We hope that this presentation has been informative and insightful and has inspired you to explore new ways of improving injury prevention and management in sports..