DEEP NEURAL NETWORK FOR CLICK-THROUGH RATE PREDICTION
Keywords:Click-Through Rate, Deep Neural Network, Preprocessing, Advertising, Social Network
Predicted clickthrough rate is one of the most frequently used criteria to determine the effectiveness of an ad. In advertising production, click-through predictions are very influential for the company that places the ad. In addition to predicting the click-through rate of an ad, the use of the model or algorithm used is also very important in analyzing the click-through rate that occurs. The purpose of this study is to compare two advertising and social network datasets, by proposing the application of the Deep Neural Network (DNN) model by testing hyperparameter variations to find a better architecture in predicting click-through rates. The hyperparameter variations include 3 variations of the hidden layer, 2 variations of the activation function, namely ReLu and Sigmoid, 3 variations of the optimizer (RMSprop, Adam, and Adagrad) and 3 variations of the learning rate (0.1, 0.01, and 0.001). Experiments conducted with the advertising parameter dataset with hidden layer 3, learning rate 0.01 and Adam optimization resulted in the highest model with an accuracy value of 99.90%, AUC 99.90% and Precision-Recall 99.89%, while the social network ads parameter data with hidden layer 5, learning rate 0.1 and Adam optimization resulted in the highest model with an accuracy value of 92.25%, AUC 92.72% and Precision-Recall 89.70%.
How to Cite
Copyright (c) 2022 University Malaysia Pahang Publishing
This work is licensed under a Creative Commons Attribution 4.0 International License.