Massive volumes of electronic health information, including data on vital signs and electrocardiograms (ECGs), are now accessible due to the big data revolution. These signals are now more easily obtained and are often captured as a time series of observations. There is a particular need to provide innovative methods that enable effective monitoring of these signals and prompt anomaly detection given the proliferation of smart devices with ECG capabilities. However, anomaly detection is still a very difficult process since the majority of created data is not yet labelled. Deep generative models have been used for unsupervised representation learning to develop expressive feature representations of sequences, which may improve the accuracy and ease of use of downstream tasks like anomaly detection. We suggest using an autoencoder to learn representations of ECG sequences in an unsupervised manner. Then, we use several detection algorithms to identify anomalies based on the acquired representations. We evaluated our method using the UCR time series classification archive's ECG5000 electrocardiogram dataset. Our findings demonstrate that the suggested strategy outperforms previous supervised and unsupervised techniques in identifying abnormalities and learning expressive representations of ECG sequences