华为云AI开发平台ModelArtsk均值_云淘科技

概述

“K-均值”节点用于产生聚类模型,用户在使用时需要指定聚类个数。K-均值算法是基于距离的算法,将所有数据归类到其最邻近的中心。

输入

参数

子参数说明

参数说明

inputs

dataframe

inputs为字典类型,dataframe为pyspark中的DataFrame类型对象

输出

spark pipeline类型的模型

参数说明

参数

子参数说明

参数说明

b_use_default_encoder

是否使用默认编码,默认为True

input_features_str

输入的列名以逗号分隔组成的字符串,例如:

“column_a”

“column_a,column_b”

cluster_feature_vector_col

算子输入的特征向量列的列名,默认为”model_features”

prediction_col

pyspark kmeans聚类器输出的预测列

k

聚类的个数,默认为2

init_mode

聚类采用的初始算法,random、k-means,默认为”random”

init_steps

采用k-means|| 初始化模式的步数,默认为2

max_iter

最大迭代次数,默认为20

tol

迭代算法的收敛阈值,默认为1e-4

样例

inputs = {
    "dataframe": None  # @input {"label":"dataframe","type":"DataFrame"}
}
params = {
    "inputs": inputs,
    "b_output_action": True,
    "b_use_default_encoder": True,  # @param {"label": "b_use_default_encoder", "type": "boolean", "required": "true", "helpTip": ""}
    "outer_pipeline_stages": None,
    "input_features_str": "",  # @param {"label": "input_features_str", "type": "string", "required": "false", "helpTip": ""}
    "cluster_feature_vector_col": "model_features",  # @param {"label": "cluster_feature_vector_col", "type": "string", "required": "true", "helpTip": ""}
    "prediction_col": "prediction",  # @param {"label": "prediction_col", "type": "string", "required": "true", "helpTip": ""}
    "k": 2,  # @param {"label": "k", "type": "integer", "required": "true", "range": "(0,2147483647]", "helpTip": ""}
    "init_mode": "random",  # @param {"label": "init_mode", "type": "string", "required": "true", "options": "random,k-means", "helpTip": ""}
    "init_steps": 2,  # @param {"label": "init_steps", "type": "integer", "required": "true", "range": "(0,2147483647]", "helpTip": ""}
    "max_iter": 20,  # @param {"label": "max_iter", "type": "integer", "required": "true", "range": "(0,2147483647]", "helpTip": ""}
    "tol": 1e-4  # @param {"label": "tol", "type": "number", "required": "true", "range": "(0.0,none)", "helpTip": ""}
}
kmeans____id___ = MLSKmeans(**params)
kmeans____id___.run()
# @output {"label":"pipeline_model","name":"kmeans____id___.get_outputs()['output_port_1']","type":"PipelineModel"}

父主题: 聚类

同意关联代理商云淘科技,购买华为云产品更优惠(QQ 78315851)

内容没看懂? 不太想学习?想快速解决? 有偿解决: 联系专家