Phần mềm-Ứng dụng

We built the neural network for data learning and data generation for the IMA application as shown in the following algorithm.

Input: Data table of user's behavior.

Output: The data set is generated by the ANN.

1. Initializing the ANN.

2. Conducting data learning.

     2.1. Opening data table.

     2.2. Transfering the values of the first record into the ANN.

     2.3  The ANN calculates with these values.

     2.4. The ANN calculates the derivatives.

     2.5. Going to the next record, go back to step 2.1, and repeat to the end of the table.

     2.6. Calculating the total derivative.

     2.7. The ANN changes its weight.

3. Using the trained ANN to compute on the test data set.

4. Using the trained ANN to generate test data for IMA.

      4.1. The ANN receives input data.

     4.2. The ANN calculates and generates the output result. This result is transfered to the IMA.

      4.3. The IMA handles the data and generates the output result.

      4.4. The model receives the results from the IMA and calculates the current state.

     4.5. From that state, ANN continues generating test data.