SymetryML Documentation
Models

Auto-select optimal model

Auto-select the optimal model for a given task (binary classifier or regression). Under the hood, SymetryML builds many different models across various permutations of the selected input attributes, and also considers various model types for the given task, then chooses the best one. The request body is an `MLContext` object.

POST
/{user}/projects/{project}/autoselect/{dsName}

Auto-select the optimal model for a given task (binary classifier or regression). Under the hood, SymetryML builds many different models across various permutations of the selected input attributes, and also considers various model types for the given task, then chooses the best one.

The request body is an MLContext object.

Authorization

SymetryMLAuth
AuthorizationBearer <token>

HMAC-SHA256 signature-based authentication: requests are signed with your secret key and sent with the Customer-ID, Sym-date, Authorization, Content-MD5 and sym-version headers. See SymetryML REST API Security for the signature algorithm and a complete example.

In: header

Path Parameters

user*string

User/Customer identifier

project*string

Project identifier

dsName*string

Data source name

Query Parameters

task*string

Machine learning model. Must match the request body example you're using below — the rendered "Try it" query string always shows this parameter's single example value, so adjust task manually.

Value in

  • "binary_classifier"
  • "multiclass_classifier"
  • "regression"
  • "anomaly"
modelid*string

Unique identifier for the model

Request Body

application/json

TypeScript Definitions

Use the request body type in TypeScript.

Response Body

application/json

application/json

application/json

application/json

curl -X POST "https://example.com/c1/projects/nbra_prj_iris/autoselect/nbra_ds_sftp_iris?task=binary_classifier&modelid=iris_autoselect_model" \  -H "Content-Type: application/json" \  -d '{    "inputAttributes": [      0,      1,      2,      3,      4,      5,      6,      7,      8,      9,      10,      11    ],    "targets": [      12    ],    "extraParameters": {      "autoselect_grid_type": "autoselect_grid_type_normal",      "selector_type": "selector_type_fw_bw",      "selector_max_iterations": "5",      "autoselect_thread_pool_size": "4",      "sml_model_assessment_type": "auc"    }  }'
{  "statusCode": 200,  "statusString": "Job is finished",  "values": {}}
{  "statusCode": 202,  "statusString": "SelectModel Job Created",  "values": {}}
{  "statusCode": 400,  "statusString": "java.lang.IllegalArgumentException: Invalid Model Name: string is null",  "values": {}}
{  "statusCode": 500,  "statusString": "",  "values": {}}

Learn frame data POST

Learns from a raw `DataFrame` posted inline (as opposed to `dss/{dsName}/learn`, which learns from a registered data source via an always-async job). Set `async=true` to run as a background job — the response is `202` with the job id in the `sym-job-id` header and an empty body; poll `GET /{user}/jobs/{jobId}` for completion. Without `async` (or `async=false`), learning happens synchronously and the `200` response body has empty `values` — it is not a job-status payload.

Auto-select optimal model (dataframe body) POST

Dataframe-body variant of auto-select: instead of a `{dsName}` path segment referencing an already-uploaded data source, the request body carries both the `dataframe` (raw data) and `mlcontext` (`inputAttributes`/`targets`) directly. Attribute resolution still goes through the project's own learned data, not the inline dataframe — the project must already have been `/learn`ed with the same schema; the body's `dataframe` supplies the scoring/validation data. Schedules an asynchronous job (same response shape/status strings as the `{dsName}` variant, including the shared `"SelectModel Job Created"` status string).