Akademik Veri Yönetim Sistemi
9th International Artificial Intelligence and Data Processing Symposium, IDAP 2025, Malatya, Türkiye, 6 - 07 Eylül 2025, (Tam Metin Bildiri)
The concept of happiness has been an important topic throughout history. The happiness of an individual has been a critical factor indicating the general well-being of society. Considering that happiness is a subjective and multidimensional concept, various scales and indices have been put forward for happiness analysis. One of these, the World Happiness Report, has provided important information to evaluate and understand various aspects of happiness. In this study, the potential of using data from the 2023 World Happiness Report in neural networkbased learning algorithms to predict the happiness index is examined. While obtaining data from this report, the data between the years 2017-2022, which included recent data, is filtered. In order to reduce the data dimensionality, the correlation matrix is calculated and the number of parameters is reduced from eight to five by applying principal component analysis. Thus, two new datasets are obtained. Standard scaling is applied to the data before being given to the models. The models are trained with the data between 2017-2021 and the happiness index is predicted for the year 2022. R2 score is calculated to verify the performance of the models and MSE value is calculated to evaluate the error rates. The RNN model gave better results in all datasets. The RNN model gave the best performance in the dataset containing all parameters with an R2 score of 0.93672 and an MSE value of 0.09361. These results show that neural network-based learning algorithms can be an effective tool in the process of predicting the happiness index in future research.