Effect of activation function and model optimizer on the performance of human activity recognition system using various deep learning models

Human Activity Recognition (HAR) plays an important role in healthcare, surveillance, and smart environments, where reliable action recognition supports timely decision-making and automation. Although deep learning-based HAR systems are widely adopted, the impact of Activation Functions (AF) and Model Optimizers (MO) on performance has not been sufficiently analyzed, especially in terms of how their combinations influence model behavior in practical scenarios. Most existing studies focus on architecture design, while the interaction between AF and MO choices remains relatively unexplored. In this work, we investigate the effect of three commonly used activation functions (ReLU, Sigmoid, and Tanh) combined with four optimization algorithms (SGD, Adam, RMSprop, and Adagrad) using two recurrent deep learning architectures, namely Bi-directional Long Short Term Memory (BiLSTM) and Convolutional Long Short-Term Memory (ConvLSTM). Experiments are conducted on six medically relevant activity classes selected from the HMDB51 and UCF101 datasets, considering their suitability for healthcare-oriented HAR applications. Our experimental results show that ConvLSTM consistently outperforms BiLSTM across both datasets. ConvLSTM combined with Adam or RMSprop achieves accuracy levels of up to 99.00%, demonstrating strong spatio-temporal learning capability and stable performance. While BiLSTM performs reasonably well on UCF101 with accuracy close to 98.00%, its performance significantly drops to around 60.00% on HMDB51, indicating limited robustness across datasets and weaker sensitivity to AF and MO variations. These findings highlight that model performance in HAR is strongly influenced by the choice of activation function and optimizer. Based on our analysis, ConvLSTM with ReLU and either Adam or RMSprop emerges as a reliable and effective configuration for accurate human activity recognition. This study provides practical insights for optimizing HAR systems, particularly for real-world healthcare environments where fast and precise activity detection is critical.