aws-sdk-sagemaker 1.175.0 → 1.176.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- checksums.yaml +4 -4
- data/CHANGELOG.md +5 -0
- data/VERSION +1 -1
- data/lib/aws-sdk-sagemaker/client.rb +3 -1
- data/lib/aws-sdk-sagemaker/client_api.rb +2 -0
- data/lib/aws-sdk-sagemaker/types.rb +34 -172
- data/lib/aws-sdk-sagemaker.rb +1 -1
- metadata +2 -2
checksums.yaml
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
---
|
2
2
|
SHA256:
|
3
|
-
metadata.gz:
|
4
|
-
data.tar.gz:
|
3
|
+
metadata.gz: b11b07add56f85c6f243d0bb5625b83fb317f8c16b23c9835046f445246624d9
|
4
|
+
data.tar.gz: 1b437a3a9d8f8e23e313a2deb12267c5ea702d648cba09d8bea942787449734f
|
5
5
|
SHA512:
|
6
|
-
metadata.gz:
|
7
|
-
data.tar.gz:
|
6
|
+
metadata.gz: c4a2d3d9cee2273b34f4c4adce3aad3e2de44b6683ee7bf89273df96865a35ad29d62ebdb808635fcc1705c13d91bc95b56875e5325e3e8d135dd3de345fef38
|
7
|
+
data.tar.gz: c91d6b9501a9718817346a1e676582a914d70e4bfb2e6a0ed55b19eadaf141e97806132e871b3405f53987bfd3021da3cfd6cdef7192cfadfb7dabebcd854134
|
data/CHANGELOG.md
CHANGED
data/VERSION
CHANGED
@@ -1 +1 @@
|
|
1
|
-
1.
|
1
|
+
1.176.0
|
@@ -1280,6 +1280,7 @@ module Aws::SageMaker
|
|
1280
1280
|
# target_attribute_name: "TargetAttributeName", # required
|
1281
1281
|
# content_type: "ContentType",
|
1282
1282
|
# channel_type: "training", # accepts training, validation
|
1283
|
+
# sample_weight_attribute_name: "SampleWeightAttributeName",
|
1283
1284
|
# },
|
1284
1285
|
# ],
|
1285
1286
|
# output_data_config: { # required
|
@@ -9929,6 +9930,7 @@ module Aws::SageMaker
|
|
9929
9930
|
# resp.input_data_config[0].target_attribute_name #=> String
|
9930
9931
|
# resp.input_data_config[0].content_type #=> String
|
9931
9932
|
# resp.input_data_config[0].channel_type #=> String, one of "training", "validation"
|
9933
|
+
# resp.input_data_config[0].sample_weight_attribute_name #=> String
|
9932
9934
|
# resp.output_data_config.kms_key_id #=> String
|
9933
9935
|
# resp.output_data_config.s3_output_path #=> String
|
9934
9936
|
# resp.role_arn #=> String
|
@@ -23470,7 +23472,7 @@ module Aws::SageMaker
|
|
23470
23472
|
params: params,
|
23471
23473
|
config: config)
|
23472
23474
|
context[:gem_name] = 'aws-sdk-sagemaker'
|
23473
|
-
context[:gem_version] = '1.
|
23475
|
+
context[:gem_version] = '1.176.0'
|
23474
23476
|
Seahorse::Client::Request.new(handlers, context)
|
23475
23477
|
end
|
23476
23478
|
|
@@ -1648,6 +1648,7 @@ module Aws::SageMaker
|
|
1648
1648
|
SageMakerImageVersionAlias = Shapes::StringShape.new(name: 'SageMakerImageVersionAlias')
|
1649
1649
|
SageMakerImageVersionAliases = Shapes::ListShape.new(name: 'SageMakerImageVersionAliases')
|
1650
1650
|
SagemakerServicecatalogStatus = Shapes::StringShape.new(name: 'SagemakerServicecatalogStatus')
|
1651
|
+
SampleWeightAttributeName = Shapes::StringShape.new(name: 'SampleWeightAttributeName')
|
1651
1652
|
SamplingPercentage = Shapes::IntegerShape.new(name: 'SamplingPercentage')
|
1652
1653
|
ScheduleConfig = Shapes::StructureShape.new(name: 'ScheduleConfig')
|
1653
1654
|
ScheduleExpression = Shapes::StringShape.new(name: 'ScheduleExpression')
|
@@ -2235,6 +2236,7 @@ module Aws::SageMaker
|
|
2235
2236
|
AutoMLChannel.add_member(:target_attribute_name, Shapes::ShapeRef.new(shape: TargetAttributeName, required: true, location_name: "TargetAttributeName"))
|
2236
2237
|
AutoMLChannel.add_member(:content_type, Shapes::ShapeRef.new(shape: ContentType, location_name: "ContentType"))
|
2237
2238
|
AutoMLChannel.add_member(:channel_type, Shapes::ShapeRef.new(shape: AutoMLChannelType, location_name: "ChannelType"))
|
2239
|
+
AutoMLChannel.add_member(:sample_weight_attribute_name, Shapes::ShapeRef.new(shape: SampleWeightAttributeName, location_name: "SampleWeightAttributeName"))
|
2238
2240
|
AutoMLChannel.struct_class = Types::AutoMLChannel
|
2239
2241
|
|
2240
2242
|
AutoMLContainerDefinition.add_member(:image, Shapes::ShapeRef.new(shape: ContainerImage, required: true, location_name: "Image"))
|
@@ -1939,6 +1939,27 @@ module Aws::SageMaker
|
|
1939
1939
|
# [1]: https://docs.aws.amazon.com/sagemaker/latest/dg/autopilot-datasets-problem-types.html#autopilot-data-sources-training-or-validation
|
1940
1940
|
# @return [String]
|
1941
1941
|
#
|
1942
|
+
# @!attribute [rw] sample_weight_attribute_name
|
1943
|
+
# If specified, this column name indicates which column of the dataset
|
1944
|
+
# should be treated as sample weights for use by the objective metric
|
1945
|
+
# during the training, evaluation, and the selection of the best
|
1946
|
+
# model. This column is not considered as a predictive feature. For
|
1947
|
+
# more information on Autopilot metrics, see [Metrics and
|
1948
|
+
# validation][1].
|
1949
|
+
#
|
1950
|
+
# Sample weights should be numeric, non-negative, with larger values
|
1951
|
+
# indicating which rows are more important than others. Data points
|
1952
|
+
# that have invalid or no weight value are excluded.
|
1953
|
+
#
|
1954
|
+
# Support for sample weights is available in [Ensembling][2] mode
|
1955
|
+
# only.
|
1956
|
+
#
|
1957
|
+
#
|
1958
|
+
#
|
1959
|
+
# [1]: https://docs.aws.amazon.com/sagemaker/latest/dg/autopilot-metrics-validation.html
|
1960
|
+
# [2]: https://docs.aws.amazon.com/sagemaker/latest/APIReference/API_AutoMLAlgorithmConfig.html
|
1961
|
+
# @return [String]
|
1962
|
+
#
|
1942
1963
|
# @see http://docs.aws.amazon.com/goto/WebAPI/sagemaker-2017-07-24/AutoMLChannel AWS API Documentation
|
1943
1964
|
#
|
1944
1965
|
class AutoMLChannel < Struct.new(
|
@@ -1946,7 +1967,8 @@ module Aws::SageMaker
|
|
1946
1967
|
:compression_type,
|
1947
1968
|
:target_attribute_name,
|
1948
1969
|
:content_type,
|
1949
|
-
:channel_type
|
1970
|
+
:channel_type,
|
1971
|
+
:sample_weight_attribute_name)
|
1950
1972
|
SENSITIVE = []
|
1951
1973
|
include Aws::Structure
|
1952
1974
|
end
|
@@ -2217,177 +2239,13 @@ module Aws::SageMaker
|
|
2217
2239
|
#
|
2218
2240
|
# @!attribute [rw] metric_name
|
2219
2241
|
# The name of the objective metric used to measure the predictive
|
2220
|
-
# quality of a machine learning system.
|
2221
|
-
#
|
2222
|
-
#
|
2223
|
-
#
|
2224
|
-
#
|
2225
|
-
#
|
2226
|
-
#
|
2227
|
-
#
|
2228
|
-
# : The ratio of the number of correctly classified items to the total
|
2229
|
-
# number of (correctly and incorrectly) classified items. It is used
|
2230
|
-
# for both binary and multiclass classification. Accuracy measures
|
2231
|
-
# how close the predicted class values are to the actual values.
|
2232
|
-
# Values for accuracy metrics vary between zero (0) and one (1). A
|
2233
|
-
# value of 1 indicates perfect accuracy, and 0 indicates perfect
|
2234
|
-
# inaccuracy.
|
2235
|
-
#
|
2236
|
-
# AUC
|
2237
|
-
#
|
2238
|
-
# : The area under the curve (AUC) metric is used to compare and
|
2239
|
-
# evaluate binary classification by algorithms that return
|
2240
|
-
# probabilities, such as logistic regression. To map the
|
2241
|
-
# probabilities into classifications, these are compared against a
|
2242
|
-
# threshold value.
|
2243
|
-
#
|
2244
|
-
# The relevant curve is the receiver operating characteristic curve
|
2245
|
-
# (ROC curve). The ROC curve plots the true positive rate (TPR) of
|
2246
|
-
# predictions (or recall) against the false positive rate (FPR) as a
|
2247
|
-
# function of the threshold value, above which a prediction is
|
2248
|
-
# considered positive. Increasing the threshold results in fewer
|
2249
|
-
# false positives, but more false negatives.
|
2250
|
-
#
|
2251
|
-
# AUC is the area under this ROC curve. Therefore, AUC provides an
|
2252
|
-
# aggregated measure of the model performance across all possible
|
2253
|
-
# classification thresholds. AUC scores vary between 0 and 1. A
|
2254
|
-
# score of 1 indicates perfect accuracy, and a score of one half
|
2255
|
-
# (0.5) indicates that the prediction is not better than a random
|
2256
|
-
# classifier.
|
2257
|
-
#
|
2258
|
-
# BalancedAccuracy
|
2259
|
-
#
|
2260
|
-
# : `BalancedAccuracy` is a metric that measures the ratio of accurate
|
2261
|
-
# predictions to all predictions. This ratio is calculated after
|
2262
|
-
# normalizing true positives (TP) and true negatives (TN) by the
|
2263
|
-
# total number of positive (P) and negative (N) values. It is used
|
2264
|
-
# in both binary and multiclass classification and is defined as
|
2265
|
-
# follows: 0.5*((TP/P)+(TN/N)), with values ranging from 0 to 1.
|
2266
|
-
# `BalancedAccuracy` gives a better measure of accuracy when the
|
2267
|
-
# number of positives or negatives differ greatly from each other in
|
2268
|
-
# an imbalanced dataset. For example, when only 1% of email is spam.
|
2269
|
-
#
|
2270
|
-
# F1
|
2271
|
-
#
|
2272
|
-
# : The `F1` score is the harmonic mean of the precision and recall,
|
2273
|
-
# defined as follows: F1 = 2 * (precision * recall) / (precision +
|
2274
|
-
# recall). It is used for binary classification into classes
|
2275
|
-
# traditionally referred to as positive and negative. Predictions
|
2276
|
-
# are said to be true when they match their actual (correct) class,
|
2277
|
-
# and false when they do not.
|
2278
|
-
#
|
2279
|
-
# Precision is the ratio of the true positive predictions to all
|
2280
|
-
# positive predictions, and it includes the false positives in a
|
2281
|
-
# dataset. Precision measures the quality of the prediction when it
|
2282
|
-
# predicts the positive class.
|
2283
|
-
#
|
2284
|
-
# Recall (or sensitivity) is the ratio of the true positive
|
2285
|
-
# predictions to all actual positive instances. Recall measures how
|
2286
|
-
# completely a model predicts the actual class members in a dataset.
|
2287
|
-
#
|
2288
|
-
# F1 scores vary between 0 and 1. A score of 1 indicates the best
|
2289
|
-
# possible performance, and 0 indicates the worst.
|
2290
|
-
#
|
2291
|
-
# F1macro
|
2292
|
-
#
|
2293
|
-
# : The `F1macro` score applies F1 scoring to multiclass
|
2294
|
-
# classification problems. It does this by calculating the precision
|
2295
|
-
# and recall, and then taking their harmonic mean to calculate the
|
2296
|
-
# F1 score for each class. Lastly, the F1macro averages the
|
2297
|
-
# individual scores to obtain the `F1macro` score. `F1macro` scores
|
2298
|
-
# vary between 0 and 1. A score of 1 indicates the best possible
|
2299
|
-
# performance, and 0 indicates the worst.
|
2300
|
-
#
|
2301
|
-
# MAE
|
2302
|
-
#
|
2303
|
-
# : The mean absolute error (MAE) is a measure of how different the
|
2304
|
-
# predicted and actual values are, when they're averaged over all
|
2305
|
-
# values. MAE is commonly used in regression analysis to understand
|
2306
|
-
# model prediction error. If there is linear regression, MAE
|
2307
|
-
# represents the average distance from a predicted line to the
|
2308
|
-
# actual value. MAE is defined as the sum of absolute errors divided
|
2309
|
-
# by the number of observations. Values range from 0 to infinity,
|
2310
|
-
# with smaller numbers indicating a better model fit to the data.
|
2311
|
-
#
|
2312
|
-
# MSE
|
2313
|
-
#
|
2314
|
-
# : The mean squared error (MSE) is the average of the squared
|
2315
|
-
# differences between the predicted and actual values. It is used
|
2316
|
-
# for regression. MSE values are always positive. The better a model
|
2317
|
-
# is at predicting the actual values, the smaller the MSE value is
|
2318
|
-
#
|
2319
|
-
# Precision
|
2320
|
-
#
|
2321
|
-
# : Precision measures how well an algorithm predicts the true
|
2322
|
-
# positives (TP) out of all of the positives that it identifies. It
|
2323
|
-
# is defined as follows: Precision = TP/(TP+FP), with values ranging
|
2324
|
-
# from zero (0) to one (1), and is used in binary classification.
|
2325
|
-
# Precision is an important metric when the cost of a false positive
|
2326
|
-
# is high. For example, the cost of a false positive is very high if
|
2327
|
-
# an airplane safety system is falsely deemed safe to fly. A false
|
2328
|
-
# positive (FP) reflects a positive prediction that is actually
|
2329
|
-
# negative in the data.
|
2330
|
-
#
|
2331
|
-
# PrecisionMacro
|
2332
|
-
#
|
2333
|
-
# : The precision macro computes precision for multiclass
|
2334
|
-
# classification problems. It does this by calculating precision for
|
2335
|
-
# each class and averaging scores to obtain precision for several
|
2336
|
-
# classes. `PrecisionMacro` scores range from zero (0) to one (1).
|
2337
|
-
# Higher scores reflect the model's ability to predict true
|
2338
|
-
# positives (TP) out of all of the positives that it identifies,
|
2339
|
-
# averaged across multiple classes.
|
2340
|
-
#
|
2341
|
-
# R2
|
2342
|
-
#
|
2343
|
-
# : R2, also known as the coefficient of determination, is used in
|
2344
|
-
# regression to quantify how much a model can explain the variance
|
2345
|
-
# of a dependent variable. Values range from one (1) to negative one
|
2346
|
-
# (-1). Higher numbers indicate a higher fraction of explained
|
2347
|
-
# variability. `R2` values close to zero (0) indicate that very
|
2348
|
-
# little of the dependent variable can be explained by the model.
|
2349
|
-
# Negative values indicate a poor fit and that the model is
|
2350
|
-
# outperformed by a constant function. For linear regression, this
|
2351
|
-
# is a horizontal line.
|
2352
|
-
#
|
2353
|
-
# Recall
|
2354
|
-
#
|
2355
|
-
# : Recall measures how well an algorithm correctly predicts all of
|
2356
|
-
# the true positives (TP) in a dataset. A true positive is a
|
2357
|
-
# positive prediction that is also an actual positive value in the
|
2358
|
-
# data. Recall is defined as follows: Recall = TP/(TP+FN), with
|
2359
|
-
# values ranging from 0 to 1. Higher scores reflect a better ability
|
2360
|
-
# of the model to predict true positives (TP) in the data, and is
|
2361
|
-
# used in binary classification.
|
2362
|
-
#
|
2363
|
-
# Recall is important when testing for cancer because it's used to
|
2364
|
-
# find all of the true positives. A false positive (FP) reflects a
|
2365
|
-
# positive prediction that is actually negative in the data. It is
|
2366
|
-
# often insufficient to measure only recall, because predicting
|
2367
|
-
# every output as a true positive yield a perfect recall score.
|
2368
|
-
#
|
2369
|
-
# RecallMacro
|
2370
|
-
#
|
2371
|
-
# : The RecallMacro computes recall for multiclass classification
|
2372
|
-
# problems by calculating recall for each class and averaging scores
|
2373
|
-
# to obtain recall for several classes. RecallMacro scores range
|
2374
|
-
# from 0 to 1. Higher scores reflect the model's ability to predict
|
2375
|
-
# true positives (TP) in a dataset. Whereas, a true positive
|
2376
|
-
# reflects a positive prediction that is also an actual positive
|
2377
|
-
# value in the data. It is often insufficient to measure only
|
2378
|
-
# recall, because predicting every output as a true positive yields
|
2379
|
-
# a perfect recall score.
|
2380
|
-
#
|
2381
|
-
# RMSE
|
2382
|
-
#
|
2383
|
-
# : Root mean squared error (RMSE) measures the square root of the
|
2384
|
-
# squared difference between predicted and actual values, and it's
|
2385
|
-
# averaged over all values. It is used in regression analysis to
|
2386
|
-
# understand model prediction error. It's an important metric to
|
2387
|
-
# indicate the presence of large model errors and outliers. Values
|
2388
|
-
# range from zero (0) to infinity, with smaller numbers indicating a
|
2389
|
-
# better model fit to the data. RMSE is dependent on scale, and
|
2390
|
-
# should not be used to compare datasets of different sizes.
|
2242
|
+
# quality of a machine learning system. During training, the model's
|
2243
|
+
# parameters are updated iteratively to optimize its performance based
|
2244
|
+
# on the feedback provided by the objective metric when evaluating the
|
2245
|
+
# model on the validation dataset.
|
2246
|
+
#
|
2247
|
+
# For the list of all available metrics supported by Autopilot, see
|
2248
|
+
# [Autopilot metrics][1].
|
2391
2249
|
#
|
2392
2250
|
# If you do not specify a metric explicitly, the default behavior is
|
2393
2251
|
# to automatically use:
|
@@ -2397,6 +2255,10 @@ module Aws::SageMaker
|
|
2397
2255
|
# * `F1`: for binary classification
|
2398
2256
|
#
|
2399
2257
|
# * `Accuracy`: for multiclass classification.
|
2258
|
+
#
|
2259
|
+
#
|
2260
|
+
#
|
2261
|
+
# [1]: https://docs.aws.amazon.com/sagemaker/latest/dg/autopilot-metrics-validation.html#autopilot-metrics
|
2400
2262
|
# @return [String]
|
2401
2263
|
#
|
2402
2264
|
# @see http://docs.aws.amazon.com/goto/WebAPI/sagemaker-2017-07-24/AutoMLJobObjective AWS API Documentation
|
data/lib/aws-sdk-sagemaker.rb
CHANGED
metadata
CHANGED
@@ -1,14 +1,14 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: aws-sdk-sagemaker
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 1.
|
4
|
+
version: 1.176.0
|
5
5
|
platform: ruby
|
6
6
|
authors:
|
7
7
|
- Amazon Web Services
|
8
8
|
autorequire:
|
9
9
|
bindir: bin
|
10
10
|
cert_chain: []
|
11
|
-
date: 2023-
|
11
|
+
date: 2023-05-02 00:00:00.000000000 Z
|
12
12
|
dependencies:
|
13
13
|
- !ruby/object:Gem::Dependency
|
14
14
|
name: aws-sdk-core
|