| Predicting Time of Death from Body Temperature Using Neural Network Methods to Solve Newton's Law of Cooling. |
| Paper ID : 1011-ISCH |
| Authors |
|
Adham Atef Nabil, Abanoub Kamel Matta, Salma Assad Shatta * Mathematical department, Faculty of Science, Helwan University, Cairo, Egypt |
| Abstract |
| Predicting the time of death from body temperature measurements is a critical challenge in forensic science, often addressed using Newton's Law of Cooling. Traditional methods rely on simple mathematical models and linear assumptions, which may not always capture the complexities of real-world scenarios. Predicting the time of death from body temperature measurements is a critical challenge in forensic science, often addressed using Newton's Law of Cooling. Traditional methods rely on simple mathematical models and linear assumptions, which may not always capture the complexities of real-world scenarios. This study explores the application of neural network methods to improve the accuracy of time-of-death predictions, addressing the limitations of conventional models. By employing deep learning techniques, we use a neural network that captures complex, non-linear relationships between body temperature data and elapsed time since death. The model is trained on historical datasets, including temperature readings and confirmed times of experiment. We evaluate the neural network's performance by comparing it with exact solutions and existing data. The results reveal that neural networks substantially improve prediction accuracy, surpassing traditional approaches. This advancement represents a promising direction for forensic analysis, potentially leading to more precise and reliable time-of-death estimates in investigative contexts. The study underscores the potential of deep learning techniques to refine forensic methodologies and offer valuable tools for the field of forensic science. |
| Keywords |
| Differential Equations, Neural Networks, Newton's Law of Cooling, Experimental Data, Time of Prediction |
| Status: Abstract Accepted (Poster Presentation) |