A one hidden layer feedforward neural network to approximate a continuous function. The network is initialized with the number of hidden nodes in the hidden layer. The network trains using parameters: trainingX, trainingY, numIterations, learningRate, batchSize. The network gives a prediction using the given test data.