Response to discussions

Running head: PROCESS IN CONDUCTING A NEURAL NETWORK PROJECT 2

PROCESS IN CONDUCTING A NEURAL NETWORK PROJECT 2

Process in Conducting a Neural Network Project

Santosh Shrestha

University of Cumberlands

Business Intelligence - ITS-531

Dr. Steve Hallman

July 15, 2020

Process in Conducting a Neural Network Project

The development process of Artificial Neural Network is similar to the structured design methodogies of the traditional decision making computer systems however there are unique phases and aspects. The development of Artificial Neural Network includes nine steps. According to Shadra et al. (2020), the data to be used for the training and testing the network is collected, organized and formatted. In the second step, the data is separated into training, validation, and testing sets. During the third step, a network architecture and structure is decided and the algorithm is selected in the fourth step. In the fifth step, the network parameter is changed and their values are initialized. The sixth step resets and restarts the training process by initializing weights and starting training and validation. The training is stopped and the network weights are freezed in the seventh step. In the final two steps, the trained network is tested and deployed for use on unknown new cases (p. 334) Thus the network gets ready for the use of stand alone system or the part of a software where the data will be presented and the output will generate the expected prediction decision.

References

Sharda, R., Delen, D., Turban, E. (2020). Deep learning. Analytics, data Science, & artificial

intelligence: Systems for decision support (pp. 334-335). NJ, Pearson.