wearable sensors

Stress is an important physiological and psychological reaction which, when prolonged, has a negative impact on the health and well-being of humans. Traditional stress assessment methods tend to be subjective and are not scalable in real-time. In this research, a lightweight machine learning framework to predict stress from easily measurable environmental and activity-related features, temperature, humidity, and steps, is introduced. The proposed method is different from the approach that depends on complex physiological signals and uses minimal non-invasive inputs that can be applied in the wearable and IoT field. The stress level is considered as the target variable, and all the features used as input variables are independent from each other to prevent data leakage. There are several machine learning and deep learning models and their performances are measured, such as LightGBM, Support vector machine (SVM), AdaBoost, Gated Recurrent Unit (GRU). Bayesian hyperparameter optimization is used to improve the performance of the model and guarantee strong generalization. Experimental results show that LightGBM outperforms the other models, giving an accuracy of 97% and a balanced precision, recall and F1 score of 95%. The results demonstrate promising predictive performance on the evaluated dataset. However, since the model was validated on a relatively small publicly available dataset, further evaluation using larger and more diverse real-world datasets is required before practical deployment.

Introduction Information and communication technologies, including wearable sensors, offer accessible and ecological ways to support psychological processes underpinning technical practice in sport, such as the regulation of attention, arousal and mental load. The aim of this exploratory study is to examine whether heart rate (HR) and physical activity (PA), recorded with a chest-worn wearable, can serve as a potential, indirect correlate of the mental load experienced during golf putting training. Two research questions were of interest: (i) Do pre-shot HR and PA differ between made and missed putts? (ii) Do pre-shot HR and PA differ following a made compared with a missed putt? Methods Seven youth golfing athletes (13–16 years old; handicap (hcp) index < 12) performed a putting training exercise in which the goal was to score five consecutive putts from 3-metre distance to the hole. HR and PA were recorded continuously and matched with putting performance using video footage. Pre-shot HR and PA were taken as the mean over the 5 s before ball contact. Because the putts are nested within sessions and athletes, the analyses combine within-athlete non-parametric tests, corrected for multiple comparisons and reported with effect sizes, with linear mixed-effects models treating athlete and session as random effects. Results Pre-shot HR and PA do not differ significantly between made and missed putts (mixed-model estimate for HR: +0.09 bpm, 95% CI [−0.46, 0.64]; p = 0.76). However, pre-shot HR is higher following a made putt and lower following a missed one (+5.08 bpm, 95% CI [4.54, 5.62]; p < 0.001), an association that persists after adjusting for the pre-shot HR of the preceding putt. PA shows the opposite trend. Discussion Because no subjective ratings of mental load, arousal or affect were collected, mental load remains one plausible interpretation of this pattern rather than an established one. Pending validation against such measures, the approach could offer coaches and athletes an exploratory starting point for conversations about how training situations are experienced, and illustrates how continuous monitoring of HR may provide a potential indirect correlate of mental load in youth golfers.