A system was developed using a bagging (bootstrap-aggregating) ensemble of neural networks to classify time-series data with imbalanced class datasets. The proposed system discussed in the paper uses a Data Balanced Bagging Ensemble of Convolutional-LSTM Neural Networks (DBBE-CLSTM) to classify time series data in a class imbalanced dataset setting. We evaluate the proposed model on the UCI Epileptic Seizure Detection dataset and show that it can achieve high performance across all evaluated metrics. It achieved a validation set accuracy of 99.23% and a validation set f1-score of 0.9809 on the UCI Epileptic Seizure Detection dataset.