Design of English Learning Effectiveness Evaluation System Based on K-Means Clustering Algorithm
Table 3
Model design steps.
Step
Functions
1
Determine the evaluation target system. The number of indicators is the number of input nodes in the BP neural network; that is, the number of neurons in the input layer of the BP neural network is 15.
2
Determine the number of layers of the BP neural network. The system adopts a three-layer network model structure with an input layer, a hidden layer, and an output layer.
3
Clarify the evaluation result. The number of nodes in the output layer is 1, which is the evaluation result of a certain student’s learning effect.
4
Standardize the evaluation target value.
5
Use random numbers (usually a number between 0 and 1) to initialize the weights and network thresholds of the network nodes, input the standardized target sample values into the network, and give the corresponding expected output.
6
In the forward propagation stage, calculate the output of each layer node and calculate the error of each layer node.
7
In the back-propagation stage, correct the weights and check whether all sample pairs have been input.
8
Calculate the error. When the total error is less than the given error, the network training ends; otherwise, go to step (6) and continue training.
9
The trained network can be used for formal evaluation.