Train model (Anomaly)
Take the output from a DSP block and train an anomaly detection model using K-means or GMM. Updates are streamed over the websocket API.
Authorizations
Path parameters
projectIdintegerRequired
Project ID
learnIdintegerRequired
Learn Block ID, use the impulse functions to retrieve the ID
Body
axesinteger[]RequiredExample:
Which axes (indexes from DSP script) to include in the training set
[0,11,22]
clusterCountintegerOptionalExample:
Number of clusters for K-means, or number of components for GMM
32
minimumConfidenceRatingnumberRequiredExample:
DEPRECATED, use "thresholds" instead. Minimum confidence rating required before tagging as anomaly
0.3
skipEmbeddingsAndMemorybooleanOptional
If set, skips creating embeddings and measuring memory (used in tests)
Responses
200
OK
application/json
Responseall of
post
POST /v1/api/{projectId}/jobs/train/anomaly/{learnId} HTTP/1.1
Host: studio.edgeimpulse.com
x-api-key: YOUR_API_KEY
Content-Type: application/json
Accept: */*
Content-Length: 345
{
"axes": [
0,
11,
22
],
"clusterCount": 32,
"minimumConfidenceRating": 0.3,
"skipEmbeddingsAndMemory": true,
"thresholds": [
{
"key": "min_score",
"description": "Score threshold",
"helpText": "Threshold score for bounding boxes. If the score for a bounding box is below this the box will be discarded.",
"suggestedValue": 1,
"suggestedValueText": "text",
"value": 0.5
}
]
}
200
OK
{
"success": true,
"error": "text",
"id": 12873488112
}
Last updated
Was this helpful?