Classifications of Yoga Poses by through Image enhancement CLF Technique

Abstract

The calming and stress-relieving effects of yoga are well known. While there are many different yoga positions, the downward dog, goddess, tree, plank, and warrior are among the most well-known. Intelligent technology, which is machine learning, can differentiate between them. As a means to a healthier life, yoga has received worldwide acclaim and praise. It's crucial that you keep your body in the right position at all times while doing a yoga pose. This work explores whether the Naive Bayes Updateable performs well for making classifications of yoga poses. The NBU has the highest accuracy, at 78.33%. LB has 70% accuracy, which is the lowest. NBU has 0.78 precision, the highest among models. LB has 0.70 precision, which is the lowest. NBU's 0.78 recall is the highest among models. The LB's recall is 0.70, which is low. NBU's 0.56 kappa is the highest among models. The LB has the lowest kappa (0.39). NBU's F-Measure is 0.78, the highest among models. The LB has a 0.70 F-Measure, the lowest number. NBU has 0.56 MCC, the highest among models. The LB has a MCC of 0.4, the lowest value. The NBU has the highest ROC, 0.83. The IBK has the lowest ROC, 0.74. LB's 0.82 PRC is the highest among models. IBK has a 0.68 ROC, the lowest PRC. This work finds that the Naïve Bayes Updateable gives best performance compare with other models.