In Peru, as in many other parts of the world, there are various social problems, such as poverty, a problem that affects many households. With the advancement of artificial intelligence algorithms, techniques can be used that can analyze the impact of socioeconomic factors and analyze the causes of poverty. In this work, an Artificial Neural Network (ANN) model will be used to understand the data, and then evaluate the importance of the variables of the socioeconomic factors of poverty, from a multidimensional approach. In the applied methodology, we start from knowing the database of the National Household Survey (ENAHO) of the year 2023, using 6 of its modules, the important one being module 200, which focuses on the characteristics of household members. This study uses the Multilayer Perceptron (MLP) model that was selected for its ability to identify complex patterns and non-linear relationships in tabular data, thus allowing to effectively capture the multiple socioeconomic factors associated with poverty. After data preprocessing, 86,183 person records will be available, with which the MLP algorithm was created and trained, obtaining excellent results with an accuracy of 98.55% with data not seen in the model training, reinforcing with a G-Mean and a Balanced Accuracy of 98.0% and 98.01% respectively. The results show that the importance of total annual expenditure per family, household members, housing (own home with title), educational level (degree of studies) and employment (PEA indicator) are socioeconomic factors that influence poverty.