oscar-defelice

df_matrix = df_rating.pivot(index='UserId', columns='MovieId', values='Rating')

oscar-defelice

class TripletLossLayer(Layer):
    """
        Layer object to minimise the triplet loss.
        Here we implement the Bayesian Personal Ranking triplet loss.
    """
    def __init__(self, **kwargs):
        super(TripletLossLayer, self).__init__(**kwargs)
    
    def bpr_triplet_loss(self, inputs):
        """

oscar-defelice

def build_embedding(df, features, emb_dim = 10, name = 'embedding_layer'):
    '''
        Define the embedding neural network to encode features in a emb_dim-dimensional vector.
        
        Parameters
        ----------
            df : pandas DataFrame
                    dataframe containing input metadata
            
            features : list of str

oscar-defelice

def build_model(n_users, n_items, emb_dim = 30):
    '''
        Define the Keras Model for training 
        
        Parameters
        ----------
        
            n_users : int
                        number of users
            

oscar-defelice

def get_posneg(df, anchor):
    """
        Given a user id anchor, it gives back the max number of triplets [anchor, positive, negative]
        available.
        Triplets are randomly shuffled to better feed the training network.
        
        Parameters
        ----------
            df : Pandas DataFrame
                    Dataframe containing ratings, having user id as rows, movie id as columns

oscar-defelice

print("Starting training process!")
print("-------------------------------------")
t_start = time.time()

for i in range(1, n_iter+1):
    triplets = get_triplets_hard(batch_size, X_usr, X_item, df_matrix)
    loss = network_train.train_on_batch(triplets, None)
    n_iteration += 1
    if i % evaluate_every == 0:
        print("\n ------------- \n")

oscar-defelice

def get_triplets_hard(batch_size, X_usr, X_item, df, return_cache = False):
    """
        Returns the list of three arrays to feed the model.
        
        Parameters
        ----------
            batch_size : int
                            size of the batch.
            
            X_usr : numpy array of shape (n_users, n_user_features)

oscar-defelice

import pandas as pd

train_data = pd.read_csv('train.csv', engine='python', encoding='utf-8', header = None, names=['Class Index',  'Title',  'Text'])
test_data = pd.read_csv('test.csv', engine='python', encoding='utf-8', header = None, names=['Class Index', 'Title',  'Text'])

oscar-defelice

from keras.preprocessing.text import Tokenizer
from keras.preprocessing.sequence import pad_sequences

tokeniser = Tokenizer()
tokeniser.fit_on_texts(train_data['Text'])
tokenised_text = tokeniser.texts_to_sequences(train_data['Text'])
tokenised_text = pad_sequences(tokenised_text, maxlen=max_len)

oscar-defelice

encoded_labels = preprocessing.LabelEncoder()
y = encoded_labels.fit_transform(train_data['label'])
y = to_categorical(y)