Abstract:
"
Advertising has become an integral part of our daily lives in the modern world, but the issue of presenting irrelevant commercials to the public poses significant challenges. This problem negatively impacts both viewers and advertising organizations. When audiences are disinterested in the products or services being advertised, it results in a waste of advertising resources and damage to the company's reputation. Therefore, ensuring that advertisements are relevant to the audience and shown to the right people at the right time is crucial.
One potential solution to the problem of irrelevant ads is the utilization of machine learning algorithms, particularly Convolutional Neural Networks (CNNs). By employing a CNN model, ads can be precisely targeted to specific demographics, such as gender and age groups, which are most likely to constitute the intended audience. This approach helps prevent adverse effects for both viewers and advertisers.
The CNN algorithm can analyze vast amounts of data, including demographic information, to make accurate predictions about the most suitable advertisements for a particular audience. This strategy involves training the model multiple times and assessing its test accuracy. In this study, two models achieved very high levels of test accuracy, with the gender model averaging a 91% accuracy score and the age model averaging 90%.
These results underscore the effectiveness of using CNN-based algorithms for image categorization in advertising, demonstrating that this approach significantly improves ad relevance and engagement with the target audience."
