MOOC Dropout Prediction Using a Hybrid Algorithm Based on Decision Tree and Extreme Learning Machine

Algorithm 2

Improved ELM.

1: Give the training set , activation

function , number of hidden neurons .

2: Randomly assign input weight vector and the bias

except the connectionless weights and biases between

input layer and hidden layer with zero.

3: Calculate the hidden layer output matrix .

4: Calculate the output weight vector where is

the Moore-Penrose generalized inverse of matrix .

5: Obtain the predicted values based on the input variables.