You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
title: How to run Multivariate Anomaly Detection API (GA version) in Postman?
3
+
titleSuffix: Azure Cognitive Services
4
+
description: Learn how to detect anomalies in your data either as a batch, or on streaming data.
5
+
services: cognitive-services
6
+
author: mrbullwinkle
7
+
manager: nitinme
8
+
ms.service: cognitive-services
9
+
ms.subservice: anomaly-detector
10
+
ms.topic: how-to
11
+
ms.date: 10/01/2019
12
+
ms.author: mbullwin
13
+
---
14
+
15
+
# How to run Multivariate Anomaly Detection API (GA version) in Postman?
16
+
17
+
Click this button to fork the API collection in Postman and follow the steps below to test.
18
+
19
+
[](https://app.getpostman.com/run-collection/18763802-b90da6d8-0f98-4200-976f-546342abcade?action=collection%2Ffork&collection-url=entityId%3D18763802-b90da6d8-0f98-4200-976f-546342abcade%26entityType%3Dcollection%26workspaceId%3De1370b45-5076-4885-884f-e9a97136ddbc#?env%5BMVAD%5D=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)
20
+
21
+
1. Select environment as **MVAD**.
22
+
23
+
24
+
25
+
1. Select **Environment**, paste your Anomaly Detector `endpoint`, `key` and dataSource `url` in to the **CURRENT VALUE** column, click **Save** to let the variables take effect.
26
+
27
+
28
+
2. Select **Collections**, and click on the first API - **Create and train a model**, then click **Send**.
29
+
30
+
**Note:** If your data is one CSV file, please set the dataSchema as **OneTable**, if your data is multiple CSV files in a folder, please set the dataSchema as **MultiTable.**
31
+
32
+
33
+
34
+
35
+
3. In the response of the first API, copy the modelId and paste it in the `modelId` in **Environments**, click **Save**. Then go to **Collections**, click on the second API - **Get model status**, and click **Send**.
36
+
37
+
38
+
4. Select the third API - **Batch Detection**, and click **Send**. This API will trigger a synchronous inference task, and you should use the Get batch detection results API several times to get the status and the final results.
39
+
40
+
41
+
5. In the response of the third API, copy the resultId and paste it in the `resultId` in **Environments**, click **Save**. Then go to **Collections**, click on the fourth API - Get batch detection results, and click **Send**.
42
+
43
+
44
+
6. For the rest of the APIs, click on each and click Send to test on their request and response.
45
+
46
+
47
+
48
+
## Next Steps
49
+
*[Create an Anomaly Detector resource](create-resource.md)
50
+
*[Quickstart: Detect anomalies in your time series data using the Anomaly Detector](../quickstarts/client-libraries.md)
# Best practices for using the Multivariate Anomaly Detector API
17
+
# Best practices for using the Multivariate Anomaly Detection API
18
18
19
-
This article will provide guidance around recommended practices to follow when using the multivariate Anomaly Detector (MVAD) APIs.
19
+
This article will provide guidance around recommended practices to follow when using the multivariate Anomaly Detection (MVAD) APIs.
20
20
In this tutorial, you'll:
21
21
22
22
> [!div class="checklist"]
@@ -46,7 +46,7 @@ Now you're able to run your code with MVAD APIs without any error. What could be
46
46
47
47
### Data quantity
48
48
49
-
* The underlying model of MVAD has millions of parameters. It needs a minimum number of data points to learn an optimal set of parameters. The empirical rule is that you need to provide **15,000 or more data points (timestamps) per variable** to train the model for good accuracy. In general, the more the training data, better the accuracy. However, in cases when you're not able to accrue that much data, we still encourage you to experiment with less data and see if the compromised accuracy is still acceptable.
49
+
* The underlying model of MVAD has millions of parameters. It needs a minimum number of data points to learn an optimal set of parameters. The empirical rule is that you need to provide **5,000 or more data points (timestamps) per variable** to train the model for good accuracy. In general, the more the training data, better the accuracy. However, in cases when you're not able to accrue that much data, we still encourage you to experiment with less data and see if the compromised accuracy is still acceptable.
50
50
* Every time when you call the inference API, you need to ensure that the source data file contains just enough data points. That is normally `slidingWindow` + number of data points that **really** need inference results. For example, in a streaming case when every time you want to inference on **ONE** new timestamp, the data file could contain only the leading `slidingWindow` plus **ONE** data point; then you could move on and create another zip file with the same number of data points (`slidingWindow` + 1) but moving ONE step to the "right" side and submit for another inference job.
51
51
52
52
Anything beyond that or "before" the leading sliding window won't impact the inference result at all and may only cause performance downgrade.Anything below that may lead to an `NotEnoughInput` error.
@@ -135,15 +135,14 @@ There are some limitations in both the training and inference APIs, you should b
135
135
136
136
#### General Limitations
137
137
* Sliding window: 28-2880 timestamps, default is 300. For periodic data, set the length of 2-4 cycles as the sliding window.
138
-
* API calls: At most 20 API calls per minute.
139
-
* Variable numbers: For training and asynchronized inference, at most 301 variables.
138
+
* Variable numbers: For training and batch inference, at most 301 variables.
140
139
#### Training Limitations
141
-
* Timestamps: At most 1000000. Too few timestamps may decrease model quality. Recommend having more than 15000 timestamps.
140
+
* Timestamps: At most 1000000. Too few timestamps may decrease model quality. Recommend having more than 5,000 timestamps.
142
141
* Granularity: The minimum granularity is `per_second`.
143
142
144
-
#### Asynchronized inference limitations
143
+
#### Batch inference limitations
145
144
* Timestamps: At most 20000, at least 1 sliding window length.
146
-
#### Synchronized inference limitations
145
+
#### Streaming inference limitations
147
146
* Timestamps: At most 2880, at least 1 sliding window length.
148
147
* Detecting timestamps: From 1 to 10.
149
148
@@ -181,12 +180,6 @@ Let's use two examples to learn how MVAD's sliding window works. Suppose you hav
181
180
***Batch scenario**: You have multiple target data points to predict. Your `endTime` will be greater than your `startTime`. Inference in such scenarios is performed in a "moving window" manner. For example, MVAD will use data from `2021-01-01T00:00:00Z` to `2021-01-01T23:59:00Z` (inclusive) to determine whether data at `2021-01-02T00:00:00Z` is anomalous. Then it moves forward and uses data from `2021-01-01T00:01:00Z` to `2021-01-02T00:00:00Z` (inclusive)
182
181
to determine whether data at `2021-01-02T00:01:00Z` is anomalous. It moves on in the same manner (taking 1,440 data points to compare) until the last timestamp specified by `endTime` (or the actual latest timestamp). Therefore, your inference data source must contain data starting from `startTime` - `slidingWindow` and ideally contains in total of size `slidingWindow` + (`endTime` - `startTime`).
183
182
184
-
### Why does the service only accept zip files for training and inference when sending data asynchronously?
185
-
186
-
We use zip files because in batch scenarios, we expect the size of both training and inference data would be very large and can't be put in the HTTP request body. This allows users to perform batch inference on historical data either for model validation or data analysis.
187
-
188
-
However, this might be somewhat inconvenient for streaming inference and for high frequency data. We have a plan to add a new API specifically designed for streaming inference that users can pass data in the request body.
189
-
190
183
### What's the difference between `severity` and `score`?
191
184
192
185
Normally we recommend you to use `severity` as the filter to sift out 'anomalies' that aren't so important to your business. Depending on your scenario and data pattern, those anomalies that are less important often have relatively lower `severity` values or standalone (discontinuous) high `severity` values like random spikes.
Copy file name to clipboardExpand all lines: articles/cognitive-services/Anomaly-Detector/faq.yml
+4Lines changed: 4 additions & 0 deletions
Original file line number
Diff line number
Diff line change
@@ -44,6 +44,10 @@ sections:
44
44
How do I configure the Anomaly Detector service to be zone-resilient?
45
45
answer: |
46
46
No customer configuration is necessary to enable zone-resiliency. Zone-resiliency for Anomaly Detector resources is available by default and managed by the service itself.
47
+
- question: |
48
+
When will I be charged for Multivariate Anomaly Detection GA version?
49
+
answer: |
50
+
Multivariate Anomaly Detection will start charging since January 2023, please stay tuned in Azure Anomaly Detector [Pricing Page](https://aka.ms/adpricing).
* Check out this AI Show video to learn more about the GA version of Multivariate Anomaly Detection: [AI Show | Multivariate Anomaly Detection is Generally Available](https://learn.microsoft.com/en-us/shows/ai-show/ep-70-the-latest-from-azure-multivariate-anomaly-detection).
29
+
18
30
### Nov 2022
19
31
20
32
* Multivariate Anomaly Detection is now a generally available feature in Anomaly Detector service, with a better user experience and better model performance. Learn more about [how to get started using the latest release of Multivariate Anomaly Detection](how-to/create-resource.md).
@@ -88,6 +100,7 @@ We've also added links to some user-generated content. Those items will be marke
88
100
89
101
## Videos
90
102
103
+
* Nov 12, 2022 AI Show: [Multivariate Anomaly Detection is GA](https://learn.microsoft.com/en-us/shows/ai-show/ep-70-the-latest-from-azure-multivariate-anomaly-detection) (Seth with Louise Han).
91
104
* May 7, 2021 [New to Anomaly Detector: Multivariate Capabilities](/shows/AI-Show/New-to-Anomaly-Detector-Multivariate-Capabilities) - AI Show on the new multivariate anomaly detection APIs with Tony Xing and Seth Juarez
92
105
* April 20, 2021 AI Show Live | Episode 11| New to Anomaly Detector: Multivariate Capabilities - AI Show live recording with Tony Xing and Seth Juarez
93
106
* May 18, 2020 [Inside Anomaly Detector](/shows/AI-Show/Inside-Anomaly-Detector) - AI Show with Qun Ying and Seth Juarez
0 commit comments