scripts/nn/examples/ncf-real-data.dml

#-------------------------------------------------------------
#
# Licensed to the Apache Software Foundation (ASF) under one
# or more contributor license agreements.  See the NOTICE file
# distributed with this work for additional information
# regarding copyright ownership.  The ASF licenses this file
# to you under the Apache License, Version 2.0 (the
# "License"); you may not use this file except in compliance
# with the License.  You may obtain a copy of the License at
#
#   http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing,
# software distributed under the License is distributed on an
# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
# KIND, either express or implied.  See the License for the
# specific language governing permissions and limitations
# under the License.
#
#-------------------------------------------------------------

# Imports
source("staging/NCF.dml") as NCF

# prepare input data

data_loc = "scripts/nn/examples/data/ml-latest-small/ml-latest-small/"

# - read user/items integer-encoded vectors
train = read(data_loc + "sampled-train.csv", format="csv", header=FALSE, sep=",");
val = read(data_loc + "sampled-test.csv", format="csv", header=FALSE, sep=",");

users_train_int_encoded = train[, 1];
items_train_int_encoded = train[, 2];
targets_train = train[, 3];

users_val_int_encoded = val[, 1];
items_val_int_encoded = val[, 2];
targets_val = val[, 3];

N = max(max(items_train_int_encoded), max(items_val_int_encoded)); # number items
M = max(max(users_train_int_encoded), max(users_val_int_encoded)); # number users

print("Done reading.");

# - create user/items matrices by applying one-hot-encoding
items_train = toOneHot(items_train_int_encoded, N);
items_val = toOneHot(items_val_int_encoded, N);
users_train = toOneHot(users_train_int_encoded, M);
users_val = toOneHot(users_val_int_encoded, M);

print("Done encoding.");

# Train

epochs = 20;
batch_size = 16;

# layer dimensions
E = 8; # embedding
D1 = 64; # dense layer 1
D2 = 32; # dense layer 2
D3 = 16; # dense layer 3

[biases, weights] = NCF::train(users_train, items_train, targets_train, users_val, items_val, targets_val, epochs, batch_size, E, D1, D2, D3);