-
-
Screenshot of the Application working
Inspiration
The Launch Angle Predictor project was inspired by the growing importance of data analytics in baseball, particularly the role of launch angle in determining hit outcomes.
What it does
The Launch Angle Predictor is a data-driven tool designed to analyze and predict the optimal launch angles for baseball hitters based on various input parameters like Hit Distance and Exit Velocity.
How we built it
The Launch Angle Predictor was developed using a combination of programming languages, data analysis techniques, and machine learning algorithms. Technologies Used: Python, Streamlit, Machine Learning, HTML, CSS, JavaScript
Dataset Used: 2016_mlb_homeruns.csv taken from MLB provided datasets. Summary: 1)Built launch angle predictor application in which we will give exit velocity and hit distance as input, and it will predict the launch angle as output. 2)We had taken the dataset from MLB provided datasets. 3)First, we had used machine learning and had done EDA, preprocessing and had used KNN regressor for prediction. In this dataset we have 5499 rows and 6 columns. 4)Next, we had built .html page in which we had used CSS and JavaScript and had built the Streamlit Application page. 5)Next, we had used streamlit for backend to combine .html in which it will take exit velocity and hit distance as input, and it will give launch angle as output. The knn model will process the inputs and it give output.
Challenges we ran into
We spent significant time cleaning and preprocessing the data, implementing techniques to handle missing values and outliers. Choosing the right machine learning model was challenging due to the complexity of the relationships between input features and launch angles. Initial models did not perform as expected.
Accomplishments that we're proud of
Conducted extensive data analysis, leading to valuable insights about the relationships between exit velocity, launch angle, and hit outcomes. This analysis not only informed our model but also provided users with a deeper understanding of their performance.
What we learned
Throughout the development of the Launch Angle Predictor, i gained valuable insights and lessons that we can apply in future projects and enhance my skills.
What's next for Launch Angle Predictor
We plan to explore and implement more advanced machine learning algorithms, such as deep learning techniques, to improve the accuracy and predictive power of the model. We aim to incorporate more performance metrics, such as swing path analysis, pitch speed to provide a more comprehensive analysis of hitting performance. To increase accessibility, we are considering developing a mobile application version of the Launch Angle Predictor.
Log in or sign up for Devpost to join the conversation.