 |
International Society of Science and Applied Technologies |
| Golf Chipping Analysis Using a Feed-Forward Neural Network | ||||
| Author | Jungyeon Choi
|
|||
| Co-Author(s) | Hui Ru Ng; Jong-Hoon Youn
|
|||
| Abstract | Wearable technology has become increasingly prominent in sports analytics, enabling cost-effective and portable approaches to performance assessment. This study investigated the feasibility of using a single accelerometer sensor attached to a pitching wedge to classify golf chip shots as Good or Bad from a 25-yard distance. Eight participants, each with a unique handicap, performed 20 shots; a total of 160 shots were recorded. Linear acceleration data were collected at 100 Hz, clipped for each swing, and used to extract 16 features indicative of backswing, strike, and follow-through phases. Four machine learning algorithms (Random Forest, Logistic Regression, Decision Tree, and K-Nearest Neighbors) and a feed-forward neural network (FFNN) were trained and optimized with Python’s scikit-learn and Tensorflow, respectively. The FFNN, along with the best-performing traditional models, demonstrated relatively high classification accuracy. These findings underscore the potential for a compact, single-sensor system to provide effective shot evaluation. Future work will include larger and more diverse participant groups in real golfing conditions, aiming to improve model generalizability and facilitate real-time feedback for golf chipping.
|
|||
| Keywords | Wearable Sensor, Machine Learning, Deep Learning, Golf Chip Shot, Feed-Forward Neural Network | |||
| Article #: RQD2025-152 | ||||
Proceedings of 30th ISSAT International Conference on Reliability & Quality in Design |