Change point detection (CPD)

CDP is a service that detects sudden and abnormal shifts in time series trends. Recognizing these changes can be efficient in interpreting the causes of changes and also finding the starting point of the crisis. This service can be used in any industry that includes time series data and where monitoring sudden changes in time steps is particularly important. This service can divide data into segments based on similarity in their behaviors. Below are some of the applications:

•    Network monitoring
•   Medical condition monitoring
• Alert notification in abnormal situations
• Detecting a sudden change in a time series state in real-time  

Causality time series

Discovering causal effects between time series provides a better understanding of the influencing factors in a process. For example, inference of the causality network can help understand how physiological processes in the human body are coupled or how one region’s climate affects that of another.

This service includes a method to provide physical causality networks in scale, based on large time series with linear and non-linear dependencies. In the output graph of the service, causal relationships between time series with the number of influential lags and their strength are displayed.

Synthetic data generation

Data is a critical and fundamental requirement for any kind of AI-based task. But more often than not, the quantity or quality of data is not sufficient. Also, organizations are sometimes reluctant to make their data publicly available for analysis due to privacy concerns. In doing so, they deprive themselves of the advantages of data analysis. Synthetic data solves this problem by generating much more data initially. This service generates synthetic data by learning the statistics behind the original data in order to improve AI-based tasks for organizations by analyzing more data.

Recommender system (context-based)  

Recommender systems extract users' interests based on their behavior. This service helps users to more easily and quickly access items that match their taste based on their previous interactions. Using such a service will increase user retention and engagement and improve sales.

This service recommends text-data. For example, papers, news, and books which have been published on websites can be suggested based on semantic similarity to the text that the user is reading. Top most similar documents to the user’s intent will be recommended to them. This feature helps user retention and increases their time spent on the website.

Portfolio optimization

The goal of portfolio optimization service is to minimize risk and maximize the profit of assets portfolio by using a specific objective function. Using this service, users can define constraints to optimize their portfolio and allocate weights to optimize their assets. Numerous objective functions of this service provide global optimum.