Extract Anomaly Detection Explanation Data after use SR-CNN in Power BI

Question

Extract Anomaly Detection Explanation Data after use SR-CNN in Power BI

Ha Trang Bui 0

Hi guys,

I have struggled with finding anomalies in Power BI, I want to see which camera_id and action is an anomalies and shown in a table.

The rule to detect that: I calculated average Case/per Cam/per Day where total Cases = countrow('Data.Sample'), Average Case/Cam/Day= Total Casess/Total Camera/Total Days. Noted that: Total Camera is distincount of all camera_id from 'Data.Table'.

After that I draw a line chart which has date and average Case/Cam/Day, and turn on anomalies. The problem is that I cannot export data table which have anomaly in detail like camera_id or action.

Does anyone have solution?

User's image

Here is an Data for this: Data.Sample

Ramr-msft 17,826 Reputation points

2023-08-08T13:16:57.7333333+00:00
@Ha Trang Bui Thanks for the question, To export the data table with the anomalies in detail, you can follow these steps.

Open the Azure Time Series Insights Gen2 Explorer and curate your data**. This is the data that will be exported into Power BI**.

Once you have created a view you are satisfied with, navigate to the More actions dropdown menu and select Connect to Power BI.

Set your parameters for export:

Data format: Choose whether you want to export Aggregate data or Raw events to Power BI

Time Range: Choose whether you'd like to see a fixed time range or the latest data in Power BI.

Store Type: Choose whether you'd like to run your selected query against Warm Store or Cold Store.

After exporting the data to Power BI, you can create a table visualization that includes the camera_id and action columns. Then, you can add a filter to the table visualization to only show rows where the anomaly column is set to true.

Your answer

Ramr-msft 17,826 Reputation points

2023-08-08T13:16:57.7333333+00:00

@Ha Trang Bui Thanks for the question, To export the data table with the anomalies in detail, you can follow these steps.

Open the Azure Time Series Insights Gen2 Explorer and curate your data**. This is the data that will be exported into Power BI**.

Once you have created a view you are satisfied with, navigate to the More actions dropdown menu and select Connect to Power BI.

Set your parameters for export:

Data format: Choose whether you want to export Aggregate data or Raw events to Power BI

Time Range: Choose whether you'd like to see a fixed time range or the latest data in Power BI.

Store Type: Choose whether you'd like to run your selected query against Warm Store or Cold Store.

After exporting the data to Power BI, you can create a table visualization that includes the camera_id and action columns. Then, you can add a filter to the table visualization to only show rows where the anomaly column is set to true.

Share via

Extract Anomaly Detection Explanation Data after use SR-CNN in Power BI

Your answer