genestboost is an ML boosting library that separates the modeling algorithm from the boosting algorithm. The result is that you can boost any generic regression model, not just trees. Build a forward-thinking (forward-propagating) neural network if you wish, or build an ensemble of support vector machines if you would so desire. Mix and match link and loss functions at will.