The rapid advancements in electric vehicle technology have elevated the lithium-ion battery to the forefront as the paramount energy storage solution. The battery’s health tends to deteriorate gradually as it ages. Due to the inevitable physiochemical reactions that take place inside the battery, it undergoes degradation and at a certain point, it becomes unserviceable. The battery degradation can be estimated using state of health (SOH). This paper employs data-driven techniques to estimate the state of health (SOH) of a battery. To estimate health parameters, a vast quantity of data, such as voltage, current, and temperature, is gathered from the NASA Prognostics Center of Excellence. The data is resampled using the superior Fourier Resampling method and then fed to a machine-learning algorithm. In this study, SOH estimation is carried out using three different machine-learning techniques i.e. Long Short Term Memory (LSTM), Deep Neural Networks (DNN), and Gated Recurrent Unit (GRU). However, the performance and accuracy of SOH estimation using these algorithms are highly dependent on hyperparameter tuning. Therefore, the optimal hyperparameter tuning has been adopted in the present work to reduce the time and complexity of the estimation. Further, the performance of various proposed techniques has been compared against each other using different performance indices such as root mean square error (RMSE), mean absolute error (MAE), and R-square error. GRU technique proved to be excelling with RMSE of 0.003, MAE of 0.003, and R-square error of 0.004 while estimating the SOH of various samples of batteries. This detailed analysis will be helpful for users to evaluate the performance of a battery and plan for maintenance accurately and effectively with minimum downtime.