Application of Hyper parameter Optimization Algorithms using Big Data

In Data analytics algorithms Hyper parameters plays a crucial role and these hyper parameters are independent on the data of itself. The identification of Hyper parameters is very difficult and time consuming task in data analytics and it cannot be performed without using heuristic method. In this paper we are using hyper parameter selection algorithms such as Bayesian Search, Simulated Annealing, Tree-structured Parzen Estimators, Basin Hopping on Spark etc. The performance of the algorithms is measured by comparing the results of the algorithm with the Random search algorithm result..