Boosting Learning Machines with Function Compositions to Avoid Local Minima in Regression Problems
Pablo Zegers, Gonzalo Correa · IEEE International Conference on Neural Networks/IEEE ... International Conference on Neural Networks · 2007
We present an improved cascaded learning method that is based on mathematical functions compositions, instead of additive models as normally done in boosting approaches. Regression experiments are done to support the usefulness of this architecture and training procedure. The method allows to produce a strong learner with increased probability of avoiding local minima.