Naïve Bayes (나이브 배이스) Classifier - 실전 2

import numpy as np import pandas as pd data_path = 'https://github.com/vandit15/Movielens-Data/blob/master/ml-1m/ratings.dat?raw=true' df = pd.read_csv(data_path, header=None, sep='::', engine='python') df.columns = ['user_id', 'movie_id', 'rating', 'timestamp']

 

n_users = df['user_id'].nunique() n_movies = df['movie_id'].nunique() def load_user_rating_data(df, n_users, n_movies):     data = np.zeros([n_users, n_movies], dtype=np.intc)     movie_id_mapping = {}     for user_id, movie_id, rating in zip(df['user_id'], df['movie_id'], df['rating']):         user_id = int(user_id) - 1         if movie_id not in movie_id_mapping:             movie_id_mapping[movie_id] = len(movie_id_mapping)         data[user_id, movie_id_mapping[movie_id]] = rating     return data, movie_id_mapping data, movie_id_mapping = load_user_rating_data(df, n_users, n_movies) # rating별 통계계 values, counts = np.unique(data, return_counts=True) for value, count in zip(values, counts):     print(f'Number of rating {value}: {count}') # movie_id별 통계계 print(df['movie_id'].value_counts())

Number of rating 0: 21384031 Number of rating 1: 56174 Number of rating 2: 107557 Number of rating 3: 261197 Number of rating 4: 348971 Number of rating 5: 226310 2858   3428 260     2991 1196   2990 1210   2883 480     2672 ... 3458   1 2226   1 1815   1 398     1 2909   1 Name: movie_id, Length: 3706, dtype: int64

 

movie_id = 2858을 대상으로 추천을 할지 안할지를 알아보자.
즉 데이터셋에서 movie_id = 2858  를 제거한 상태

target_movie_id = 2858 X_raw = np.delete(data, movie_id_mapping[target_movie_id], axis=1) Y_raw = data[:, movie_id_mapping[target_movie_id]] X = X_raw[Y_raw > 0] Y = Y_raw[Y_raw > 0] print('Shape of X:', X.shape) print('Shape of Y:', Y.shape)

Shape of X: (3428, 3705) Shape of Y: (3428,)

 

추천(=1), 비추천(=0)으로 이진화하기 위해 rating을 3을 기준으로 타켓을 변환

recommend = 3 Y[Y <= recommend] = 0 Y[Y > recommend] = 1 n_pos = (Y == 1).sum() n_neg = (Y == 0).sum() print(f'{n_pos} positive samples and {n_neg} negative samples.')

2853 positive samples and 575 negative samples.

 

준비된 데이터세을 8:2 로 학습용과 테스트용으로 분리

from sklearn.model_selection import train_test_split X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size=0.2, random_state=42) print(len(Y_train), len(Y_test))

2742 686

 

Naïve Bayes 트레이닝

from sklearn.naive_bayes import MultinomialNB clf = MultinomialNB(alpha=1.0, fit_prior=True) clf.fit(X_train, Y_train)

 

Accuracy(정확도) 체크

prediction_prob = clf.predict_proba(X_test) print(prediction_prob[0:10]) prediction = clf.predict(X_test) print(prediction[:10]) print(Y_test[:10]) accuracy = clf.score(X_test, Y_test) print(f'The accuracy is: {accuracy*100:.1f}%')

[[7.50487439e-23 1.00000000e+00] [1.01806208e-01 8.98193792e-01] [3.57740570e-10 1.00000000e+00] [1.00000000e+00 2.94095407e-16] [1.00000000e+00 2.49760836e-25] [7.62630220e-01 2.37369780e-01] [3.47479627e-05 9.99965252e-01] [2.66075292e-11 1.00000000e+00] [5.88493563e-10 9.99999999e-01] [9.71326867e-09 9.99999990e-01]] Predict       -> [1 1 1 1 1 1 0 1 1 1]   10개 샘플중 4개가 틀림 Y_test[:10] -> [1 1 1 0 0 0 1 1 1 1] The accuracy is: 71.6%