首页  思维导图  详情

特征列

2019-09-21 14:34:45   0  举报





仅支持查看

AI智能生成

机器学习，特征列

作者其他创作

大纲/内容

bucketized_column

bucketized_feature_column = tf.feature_column.bucketized_column(
source_column = numeric_feature_column,
boundaries = [1960, 1980, 2000]

Dense Column

numeric_column

默认数据类型 (tf.float32)

indicator_column

indicator_column = tf.feature_column.indicator_column(categorical_column)

子主题

embedding_column

embedding_column = tf.feature_column.embedding_column(
categorical_column=categorical_column,
dimension=dimension_of_embedding_vector)

子主题

embedding_dimensions = number_of_categories**0.25

Categorical Column

categorical_column_with_identity

identity_feature_column = tf.feature_column.categorical_column_with_identity(
key='my_feature_b',
num_buckets=4)

categorical_column_with_vocabulary_list

vocabulary_feature_column =
tf.feature_column.categorical_column_with_vocabulary_list(
key="a feature returned by input_fn()",
vocabulary_list=["kitchenware", "electronics", "sports"])

categorical_column_with_vocabulary_file

ocabulary_feature_column =
tf.feature_column.categorical_column_with_vocabulary_file(
key="a feature returned by input_fn()",
vocabulary_file="product_class.txt",
vocabulary_size=3)

categorical_column_with_hash_bucket

hashed_feature_column =
tf.feature_column.categorical_column_with_hash_bucket(
key = "some_feature",
hash_buckets_size = 100) # The number of categories

crossed_column

crossed_lat_lon_fc = tf.feature_column.crossed_column(
[latitude_bucket_fc, longitude_bucket_fc], 5000)