Linear Regression, Logistic Regression and SVM all have convex functions. Convex functions means that if we found local minima, that minima is global minima. I cannot practically and intuitively understand this behaviour. Everything works mathematically but intuitively I cant understand why each cost function leads us to having global minima!