als_tf

Posted 2020-11-30 kayy

# -*- coding: utf-8 -*-

import tensorflow as tf

def als_tf(user_item_matrix, feature_num, train_times, reg):
    user_num, item_num = user_item_matrix.shape
    x = tf.placeholder(tf.float32, [user_num, item_num])
    user_feature = tf.Variable(tf.random_normal([user_num, feature_num], 0, 0.1), name=‘user_feature‘)
    item_feature = tf.Variable(tf.random_normal([item_num, feature_num], 0, 0.1), name=‘item_feature‘)

    loss = 1.0 * tf.reduce_sum((tf.matmul(user_feature, item_feature, transpose_b=True) - x) ** 2) +
           1.0 * reg * ((tf.reduce_sum(user_feature ** 2) + tf.reduce_sum(item_feature ** 2)))

    global_step = tf.Variable(0, trainable=False, name=‘global_step‘)

    learning_rate_init = 0.03
    learning_rate = tf.train.exponential_decay(learning_rate_init, global_step, 10, 0.99, staircase=True)
    train_step = tf.train.AdagradOptimizer(learning_rate).minimize(loss, global_step=global_step)

    with tf.Session() as sess:
        sess.run(tf.initialize_all_variables())
        for i in range(train_times):
            train_feed_dict = {
                x: user_item_matrix
            }
            sess.run(train_step, feed_dict=train_feed_dict)
            if i % 10 == 0:
                print(i, "train_times, loss: ", sess.run(loss, feed_dict=train_feed_dict))

import numpy as np

if __name__ == ‘__main__‘:
    x = np.random.randint(0, 1, [30, 50])
    als_tf(x, 5, 100, 0.5)