Add ML commons algorthim documentation

Signed-off-by: Naarcha-AWS <naarcha@amazon.com>
This commit is contained in:
Naarcha-AWS 2022-05-23 11:29:54 -05:00
parent 5ca9327c97
commit 8042de2098
1 changed files with 220 additions and 0 deletions

View File

@ -0,0 +1,220 @@
---
layout: default
title: Supported Algorithms
has_children: false
nav_order: 100
---
# Supported Algorithms
ML Commons supports various algorithms to help train and predict ML models or test data-driven predictions without a model. This page outlines the algorithms supported by the ML Commons plugin and the API actions they support
## Common limitation
Except for the Localization algorithm, all of the following algorithms can only support retrieving 10,000 documents from an index as an input.
## K-Means
K-Means is a simple and popular unsupervised clustering ML algorithm. K-Means will randomly choose centroids, then calculate iteratively to optimize the position of the centroids until each observation belongs to the cluster with the nearest mean.
### Parameters
Parameter | Type | Description | Default Value
:--- |:--- | :--- | :---
centroids | integer | The number of clusters to group the generated data | `2`
iterations | integer | The number of iterations to perform against the data until a mean generates | `10`
distance_type | enum, such as `EUCLIDEAN`, `COSINE`, or `L1` | Type of measurement from which to measure the distance between centroids | `EUCLIDEAN`
### APIs
* [Train](https://opensearch.org/docs/latest/ml-commons-plugin/api/#train-model)
* [Predict](https://opensearch.org/docs/latest/ml-commons-plugin/api/#predict)
* [Train and predict](https://opensearch.org/docs/latest/ml-commons-plugin/api/#train-and-predict)
### Example
The following example uses the Iris Data index to train K-Means synchronously.
```json
POST /_plugins/_ml/_train/kmeans
{
"parameters": {
"centroids": 3,
"iterations": 10,
"distance_type": "COSINE"
},
"input_query": {
"_source": ["petal_length_in_cm", "petal_width_in_cm"],
"size": 10000
},
"input_index": [
"iris_data"
]
}
```
### Limitations
The training process supports multi-threads, but the thread number is less than half of CPUs.
## Linear Regression
Linear Regression maps the linear relationship between inputs and outputs. In ml-common, the linear regression algorithm is adopted from the public machine learning library [Tribuo](https://tribuo.org/), which offers multidimensional linear regression models. The model supports the linear optimizer in training, including popular approaches like Linear Decay, SQRT_DECAY, [ADA](http://chrome-extension//gphandlahdpffmccakmbngmbjnjiiahp/https://www.jmlr.org/papers/volume12/duchi11a/duchi11a.pdf), [ADAM](https://tribuo.org/learn/4.1/javadoc/org/tribuo/math/optimisers/Adam.html), and [RMS_DROP](https://tribuo.org/learn/4.1/javadoc/org/tribuo/math/optimisers/RMSProp.html).
### Parameters
Parameter | Type | Description | Default Value
:--- |:--- | :--- | :---
learningRate | Double | The rate of speed that the gradient moves during descent | 0.01
momentumFactor | Double | The medium-term from which the regressor rises or falls | 0
epsilon | Double | The criteria in which a linear model is identified | 1.00E-06
beta1 | Double | The estimated exponential decay for the moment | 0.9
beta2 | Double | The estimated exponential decay for the moment | 0.99
decayRate | Double | The rate at which the model decays exponentially | 0.9
momentumType | MomentumType | The momentum with SDG to help accelerate gradients vectors in the right directions, leading to a faster convergence | STANDARD
optimizerType | OptimizerType | The optimizer used in the model | SIMPLE_SGD
### APIs
* [Train](https://opensearch.org/docs/latest/ml-commons-plugin/api/#train-model)
* [Predict](https://opensearch.org/docs/latest/ml-commons-plugin/api/#predict)
### Example
The following example predicts based on the previously trained linear regression model.
**Request**
```json
POST _plugins/_ml/_predict/LINEAR_REGRESSION/ROZs-38Br5eVE0lTsoD9
{
"parameters": {
"target": "price"
},
"input_data": {
"column_metas": [
{
"name": "A",
"column_type": "DOUBLE"
},
{
"name": "B",
"column_type": "DOUBLE"
}
],
"rows": [
{
"values": [
{
"column_type": "DOUBLE",
"value": 3
},
{
"column_type": "DOUBLE",
"value": 5
}
]
}
]
}
}
```
**Response**
```json
{
"status": "COMPLETED",
"prediction_result": {
"column_metas": [
{
"name": "price",
"column_type": "DOUBLE"
}
],
"rows": [
{
"values": [
{
"column_type": "DOUBLE",
"value": 17.25701855310131
}
]
}
]
}
}
```
### Limitations
ML Commons only supports the linear Stochastic gradient trainer or optimizer, which can not effectively map the non-linear relationships in trained data. When used with complicated data sets, the linear Stochastic trainer might cause some convergence problems and inaccurate results.
## RCF
[Random Cut Forest](https://github.com/aws/random-cut-forest-by-aws) (RCF) is a probabilistic data structure used primarily for unsupervised anomaly detection. Its use also extends to density estimation and forecasting. OpenSearch leverages RCF for anomaly detection. ML Commons supports two new variants of RCF for different use cases:
* Batch RCF: Detect anomalies in non-time-series data
* Fixed in time (FIT) RCF: Detect anomalies in time-series data
### Parameters
#### Batch RCF
Parameter | Type | Description | Default Value
:--- |:--- | :--- | :---
number_of_trees | integer | Number of trees in the forest | 30
sample_size | integer | The same size used by the stream samplers in the forest | 256
output_after | integer | The number of points required by stream samplers before results return | 32
training_data_size | integer | The size of your training data | Data set size
anamoly_score_threshold | double | The threshold of the anomaly score | 1.0
#### Fit RCF
Parameter | Type | Description | Default Value
:--- |:--- | :--- | :---
number_of_trees | integer | Number of trees in the forest | 30
shingle_size | integer | A shingle, or consecutive sequence of the most recent records | 8
sample_size | integer | The sample size used by stream samplers in the forest | 256
output_after | integer | The number of points required by stream samplers before results return | 32
time_decay | double | The decay factor used by stream samplers in the forest | 0.0001
anomaly_rate | double | The anomaly rate | 0.005
time_field | string | (**Required**) The time filed for RCF to use as time-series data | N/A
date_format | string | The date and time formatting for the time_field field | "yyyy-MM-ddHH:mm:ss"
time_zone | string | The time zone for the time_field field | "UTC"
### APIs
* [Train](https://opensearch.org/docs/latest/ml-commons-plugin/api/#train-model)
* [Predict](https://opensearch.org/docs/latest/ml-commons-plugin/api/#predict)
* [Train and predict](https://opensearch.org/docs/latest/ml-commons-plugin/api/#train-and-predict)
### Limitations
For FIT RCF, you can train the model with historical data, and store the trained model in your index. The model will be deserialized and predict new data points when using the Predict API. But the model in the index will not be refreshed with new data, because the model is "Fixed In Time".
## Localization
Finding subset level information for aggregate data (for example, aggregated over time) that demonstrates the activity of interest (spikes, drops, changes, anomalies) is a critical insight. Localization can be applied in different scenarios, such as data exploration, root cause analysis, etc., to expose the contributors driving the activity of interest in the aggregate data.
### Parameters
Parameter | Type | Description | Default Value
:--- | :--- | :--- | :---
index_name | String | The data collection to analyze | N/A
attribute_field_names | List<String> | The fields for entity kets | N/A
aggregations | List<AggregationBuilder> | The fields and aggregation for values | N/A
time_field_name | String | The timestamp field | null
start_time | Long | The beginning of the time range | 0
end_time | Long | The end of time range | 0
min_time_interval | Long | The minimal time interval/scale for analysis | 0
num_outputs | integer | The maximum number of values from localization/slicing | 0
filter_query | Long | (Optional) Reduces the collection of data for analysis | Optional.empty()
anomaly_star | QueryBuilder | (Optional) The time from which after the data will be analyzed | Optional.empty()
### Limitations
The localization algorithm can only be executed directly. Therefore, it cannot be used with the ML Common's train and predicts APIs.