Enhancing real-time health monitoring with hybrid recurrent long short-term tyrannosaurus search for menstrual cups

Menstrual Hygiene Management is not only a health issue but also a crucial aspect of social and economic development. When women and girls have proper menstrual hygiene, it positively impacts their overall wellbeing, educational opportunities, and participation in the workforce. This study proposed the Hybrid Recurrent Long Short-term based Tyrannosaurus Search (HRLS-TS) algorithm for real-time health monitoring during menstruation. The integration of these advanced techniques offers real-time data processing and analysis, especially when combined with IoT devices. In this work, a Recurrent Neural Network is employed to predict menstrual cycle-related historical data and analyze menstrual cycle patterns, also Long Short-term Memory (LSTM) is utilized to analyze menstrual flow data and capture rapid changes and fluctuations. To enhance accuracy, an initial search-based Tyrannosaurus Optimization technique is applied. Notably, the incorporation of Tyrannosaurus Optimization ensures efficient hyperparameter tuning, enhancing the overall performance of the developed method. Experiments were conducted based on performance evaluation measures such as recall, F1score, precision, Area Under the Curve-Receiver Operating Characteristic (AUC-ROC), accuracy and specificity, Mean Absolute Percentage error (MAPE), computational time, Mean Squared Error (MSE), Mean Absolute Error (MAE) and residual error, were used to assess the proposed and existing methods. The results are then compared with existing methods, demonstrating the efficiency of the HRLS-TS technique in real-time health monitoring.