From data to hope: deep neural network-based prediction of poisoning (DNNPPS) suicide cases

Background: Suicide is a pressing issue in modern society, causing significant societal and economic consequences. To address and minimize the negative impacts, it is crucial to implement effective prevention strategies. With the rising popularity of deep neural network (DNN) algorithms, they have been utilized in various health sectors. In our research, we examined the effectiveness of DNNs in predicting suicide cases.

Methods: In our study, we conducted a descriptive-analytical, cross-sectional investigation using a DNN algorithm with a sequential model consisting of four dense layers to analyze suicide data (DNNPPS). In this study, we utilized a suicide dataset comprising 1,500 data points, which were provided by a health research center.

Results: Factors such as history of psychiatric hospitals, days of the week, and job were identified as the most important risk factors for predicting suicide attempts. We obtained promising results by applying the DNNPPS algorithm to a dataset of 1453 individuals with a history of suicide. We approached the problem as a binary classification task, with suicide history as the target variable. We performed preprocessing techniques, including class balancing, and constructed a DNN model using a sequential architecture with four dense layers.

Conclusion: It is important to note that the success of the DNN algorithm relies on the quality and quantity of the available data, as well as the model’s architecture. For example, high-quality data should be accurate, representative, and relevant to the problem. In terms of quantity large dataset allows the DNN to learn more features. Nonetheless, in our study, the overall performance of the DNNPPS algorithm was satisfactory. For example, the f1-score value reached 91%, which indicates that the model has a high level of accuracy in predicting the positive class (in this case, suicide cases) and maintains a good balance between precision and recall.