HARJONO, CHRISTOPHER EVAN (2025) PREDICTIVE ANALYTICS FOR ELECTRONIC RETAIL USING NEURAL NETWORK BACKPROPAGATION. S1 thesis, UNIVERSITAS KATOLIK SOEGIJAPRANATA.

Preview	Text 18.K1.0078_CHRISTOPHER EVAN HARJONO_COVER_1.pdf Download (631kB) | Preview
	Text 18.K1.0078_CHRISTOPHER EVAN HARJONO_BAB I.pdf Restricted to Registered users only Download (507kB)
	Text 18.K1.0078_CHRISTOPHER EVAN HARJONO_BAB II.pdf Restricted to Registered users only Download (483kB)
	Text 18.K1.0078_CHRISTOPHER EVAN HARJONO_BAB III.pdf Restricted to Registered users only Download (1MB)
	Text 18.K1.0078_CHRISTOPHER EVAN HARJONO_BAB IV.pdf Restricted to Registered users only Download (1MB)
	Text 18.K1.0078_CHRISTOPHER EVAN HARJONO_BAB V.pdf Restricted to Registered users only Download (471kB)
Preview	Text 18.K1.0078_CHRISTOPHER EVAN HARJONO_DAFPUS.pdf Download (539kB) | Preview
	Text 18.K1.0078_CHRISTOPHER EVAN HARJONO_LAMP.pdf Restricted to Registered users only Download (810kB)

Abstract

The difficulty of selling electronic goods in determining the buy price, selling price, and negotiation price to get maximum profit is because it is done manually. With Artificial Intelligence, several algorithm models can be applied, one of which is the backpropagation neural network. Neural Network backpropagation is a model used to train neural networks with error prediction based on bias and weights. The result of the research is to predict the selling price, buying price, and negotiated price with an accuracy of 88.9% and the percentage of MSE values of the predicted buying price, selling price and negotiation price are 1.41%, 1.97%, and 1. 83%, where the total MSE value of the three variables is below 5% and can predict the selling price, buying price, and negotiation price in real time with an MSE buy value of 2.47%, an MSE sell value of 3.31%, and an MSE negotiable value of 2.36% where the total MSE value of the three variables is below 5% and the accuracy obtained is 79.67%.

Actions (login required)

View Item