Time Series

Machine learning algorithms specifically tailored to extract patterns from time series data need to deal with the time component of the data. A time series is a sequence of observations taken sequentially in time.

To handle the time component of the data, we have a powerful pipeline that can manipulate the data, in terms of forecasting or predicting future occurrences, extract time dependencies between the data, or smooth the signal when there is noise present.

Ekkono’s incremental learning algorithms handle streams of data, such as time series. Those and other batch trained algorithms can extract trends, seasonal patterns, and detect noisy data.

All of this is available both for standard CPUs and microcontrollers. 

Related content

More Research from Ekkono

Federated Learning

Federated Learning

Federated Learning is a machine learning technique to train a global model from different clients or devices while keeping the data decentralized.  A central server is in charge of coordinating the transfer of the models’ data between the server and the clients,...

Incremental Learning

Incremental Learning

Incremental Learning is a key feature when dealing with tiny and embedded devices. Many have claimed that only inference can be performed on embedded devices. At Ekkono, we have built a library that can train and run machine learning algorithms not only on big...

Change Detection

Change Detection

On edge machine learning we are constantly dealing with non-stationary environments. These non-stationary environments can range from changes in temperature or humidity in sensors to changes in the way a machine, such as a gas turbine, is operating. Time dependent...