Trauining dataset after DBSCAN clustering algorithm possible? And if yes, how?

Question

I am working on step detection algorithm, i have already extracted some features like std,mean,var, fir filtra