Accelerator Synopsis
Anomaly Detection & Hypothesis Generation aims to identify opportunities to improve and adapt to any anomalies within a specific time. Anomaly Detection & Hypothesis Generation built on state-of-the-art, proprietary AI/ML algorithms uses cloud resources to scale up anomaly detection on Big Data.
The users can create and update analytical datasets, toolkits, identify anomalies, and even create human-readable statistical explanations and hypotheses to take corrective actions.
Value Addition
Statistical Explanation
Our proprietary algorithms developed in-house can detect anomalies on a broader range of data sets and provide statistical explanations about why data points are classified as anomalies.
Hypothesis Generation
Using cutting-edge algorithms, the accelerator can identify anomalies and generate hypotheses based on other variables in the data to explain the anomalies.
Features
Datasets & Connections
Users can connect to multiple data sources, update existing connections, and upload files directly into the application. Users can choose to run the algorithms on all or a subset of their data.
Anomaly Detection
A summary page that highlights anomalies and takes user feedback into account. Further, the deep-dive option provides statistical explanations and redirects to the generation of hypotheses.
Hypothesis Generation
A single landing summary page. It highlights the hypotheses for all the anomalies and tags them as a priority for the user to look into.
Summary of Results
A landing page containing a summary of model results and critical insights from the past week and some high-level KPIs and data health indicators customized according to business needs.
Sparking AI Innovation at the Intersection of Business Analytics, Data Science, and Engineering.
FAQs
The application is based on a cloud framework and can be quickly scaled to support Big Data processing.
Yes, the advantage of this algorithm is that it can efficiently run on different types of data.