数据分析作业四-基于用户及物品数据进行内容推荐

news2025/2/4 3:59:16 
## 导入支持库
import pandas as pd
import matplotlib.pyplot as plt
import sklearn.metrics as metrics
import numpy as np
from sklearn.neighbors import NearestNeighbors
from scipy.spatial.distance import correlation
from sklearn.metrics.pairwise import pairwise_distances
import ipywidgets as widgets
from IPython.display import display, clear_output
from contextlib import contextmanager
import warnings
warnings.filterwarnings('ignore')
import numpy as np
import os, sys
import re
import seaborn as sns

## 加载数据集并检查书籍，用户和评级数据集的形状
books = pd.read_csv('F:\\data\\bleeding_data\\BX-Books.csv',
                    sep=None,encoding="latin-1")
books.columns = ['ISBN', 'bookTitle', 'bookAuthor',
                 'yearOfPublication', 'publisher',
                 'imageUrlS', 'imageUrlM', 'imageUrlL']

users = pd.read_csv('F:\\data\\bleeding_data\\BX-Users.csv',
                    sep=None, encoding="latin-1")
users.columns = ['userID', 'Location', 'Age']


ratings = pd.read_csv('F:\\data\\bleeding_data\\BX-Book-Ratings.csv',
                      sep=None, encoding="latin-1")
ratings.columns = ['userID', 'ISBN', 'bookRating']

print (books.shape)
print (users.shape)
print (ratings.shape)

(271360, 8)
(278858, 3)
(1149780, 3)

## 一、图书数据集
books.head()
ISBNbookTitlebookAuthoryearOfPublicationpublisherimageUrlSimageUrlMimageUrlL
00195153448Classical MythologyMark P. O. Morford2002Oxford University Presshttp://images.amazon.com/images/P/0195153448.0...http://images.amazon.com/images/P/0195153448.0...http://images.amazon.com/images/P/0195153448.0...
10002005018Clara CallanRichard Bruce Wright2001HarperFlamingo Canadahttp://images.amazon.com/images/P/0002005018.0...http://images.amazon.com/images/P/0002005018.0...http://images.amazon.com/images/P/0002005018.0...
20060973129Decision in NormandyCarlo D'Este1991HarperPerennialhttp://images.amazon.com/images/P/0060973129.0...http://images.amazon.com/images/P/0060973129.0...http://images.amazon.com/images/P/0060973129.0...
30374157065Flu: The Story of the Great Influenza Pandemic...Gina Bari Kolata1999Farrar Straus Girouxhttp://images.amazon.com/images/P/0374157065.0...http://images.amazon.com/images/P/0374157065.0...http://images.amazon.com/images/P/0374157065.0...
40393045218The Mummies of UrumchiE. J. W. Barber1999W. W. Norton & Companyhttp://images.amazon.com/images/P/0393045218.0...http://images.amazon.com/images/P/0393045218.0...http://images.amazon.com/images/P/0393045218.0...
## url不需要分析，进行删除
books.drop(['imageUrlS', 'imageUrlM', 'imageUrlL'],axis=1,inplace=True)
books.head()
ISBNbookTitlebookAuthoryearOfPublicationpublisher
00195153448Classical MythologyMark P. O. Morford2002Oxford University Press
10002005018Clara CallanRichard Bruce Wright2001HarperFlamingo Canada
20060973129Decision in NormandyCarlo D'Este1991HarperPerennial
30374157065Flu: The Story of the Great Influenza Pandemic...Gina Bari Kolata1999Farrar Straus Giroux
40393045218The Mummies of UrumchiE. J. W. Barber1999W. W. Norton & Company
## books.dtypes
books.dtypes

ISBN                 object
bookTitle            object
bookAuthor           object
yearOfPublication    object
publisher            object
dtype: object

## 现在检查属性的唯一值
books.bookTitle.unique()

array(['Classical Mythology', 'Clara Callan', 'Decision in Normandy', ...,
       'Lily Dale : The True Story of the Town that Talks to the Dead',
       "Republic (World's Classics)",
       "A Guided Tour of Rene Descartes' Meditations on First Philosophy with Complete Translations of the Meditations by Ronald Rubin"],
      dtype=object)

books.yearOfPublication.unique()

array(['2002', '2001', '1991', '1999', '2000', '1993', '1996', '1988',
       '2004', '1998', '1994', '2003', '1997', '1983', '1979', '1995',
       '1982', '1985', '1992', '1986', '1978', '1980', '1952', '1987',
       '1990', '1981', '1989', '1984', '0', '1968', '1961', '1958',
       '1974', '1976', '1971', '1977', '1975', '1965', '1941', '1970',
       '1962', '1973', '1972', '1960', '1966', '1920', '1956', '1959',
       '1953', '1951', '1942', '1963', '1964', '1969', '1954', '1950',
       '1967', '2005', '1957', '1940', '1937', '1955', '1946', '1936',
       '1930', '2011', '1925', '1948', '1943', '1947', '1945', '1923',
       '2020', '1939', '1926', '1938', '2030', '1911', '1904', '1949',
       '1932', '1928', '1929', '1927', '1931', '1914', '2050', '1934',
       '1910', '1933', '1902', '1924', '1921', '1900', '2038', '2026',
       '1944', '1917', '1901', '2010', '1908', '1906', '1935', '1806',
       '2021', '2012', '2006', 'DK Publishing Inc', 'Gallimard', '1909',
       '2008', '1378', '1919', '1922', '1897', '2024', '1376', '2037'],
      dtype=object)

books.loc[books.yearOfPublication == 'DK Publishing Inc',:]
books.yearOfPublication.unique()

array(['2002', '2001', '1991', '1999', '2000', '1993', '1996', '1988',
       '2004', '1998', '1994', '2003', '1997', '1983', '1979', '1995',
       '1982', '1985', '1992', '1986', '1978', '1980', '1952', '1987',
       '1990', '1981', '1989', '1984', '0', '1968', '1961', '1958',
       '1974', '1976', '1971', '1977', '1975', '1965', '1941', '1970',
       '1962', '1973', '1972', '1960', '1966', '1920', '1956', '1959',
       '1953', '1951', '1942', '1963', '1964', '1969', '1954', '1950',
       '1967', '2005', '1957', '1940', '1937', '1955', '1946', '1936',
       '1930', '2011', '1925', '1948', '1943', '1947', '1945', '1923',
       '2020', '1939', '1926', '1938', '2030', '1911', '1904', '1949',
       '1932', '1928', '1929', '1927', '1931', '1914', '2050', '1934',
       '1910', '1933', '1902', '1924', '1921', '1900', '2038', '2026',
       '1944', '1917', '1901', '2010', '1908', '1906', '1935', '1806',
       '2021', '2012', '2006', 'DK Publishing Inc', 'Gallimard', '1909',
       '2008', '1378', '1919', '1922', '1897', '2024', '1376', '2037'],
      dtype=object)

print(books.loc[books.yearOfPublication == 'DK Publishing Inc',:])

              ISBN                                          bookTitle  \
209538  078946697X  DK Readers: Creating the X-Men, How It All Beg...   
221678  0789466953  DK Readers: Creating the X-Men, How Comic Book...   

       bookAuthor  yearOfPublication  \
209538       2000  DK Publishing Inc   
221678       2000  DK Publishing Inc   

                                                publisher  
209538  http://images.amazon.com/images/P/078946697X.0...  
221678  http://images.amazon.com/images/P/0789466953.0...  

books.loc[books.yearOfPublication == 'DK Publishing Inc',:]
ISBNbookTitlebookAuthoryearOfPublicationpublisher
209538078946697XDK Readers: Creating the X-Men, How It All Beg...2000DK Publishing Inchttp://images.amazon.com/images/P/078946697X.0...
2216780789466953DK Readers: Creating the X-Men, How Comic Book...2000DK Publishing Inchttp://images.amazon.com/images/P/0789466953.0...
## 从上面可以看出，bookAuthor错误地装载了bookTitle，因此需要进行修正。
# ISBN '0789466953'
books.loc[books.ISBN == '0789466953','yearOfPublication'] = 2000
books.loc[books.ISBN == '0789466953','bookAuthor'] = "James Buckley"
books.loc[books.ISBN == '0789466953','publisher'] = "DK Publishing Inc"
books.loc[books.ISBN == '0789466953','bookTitle'] = "DK Readers: Creating the X-Men, How Comic Books Come to Life (Level 4: Proficient Readers)"

#ISBN '078946697X'
books.loc[books.ISBN == '078946697X','yearOfPublication'] = 2000
books.loc[books.ISBN == '078946697X','bookAuthor'] = "Michael Teitelbaum"
books.loc[books.ISBN == '078946697X','publisher'] = "DK Publishing Inc"
books.loc[books.ISBN == '078946697X','bookTitle'] = "DK Readers: Creating the X-Men, How It All Began (Level 4: Proficient Readers)"

books.loc[(books.ISBN == '0789466953') | (books.ISBN == '078946697X'),:]
ISBNbookTitlebookAuthoryearOfPublicationpublisher
209538078946697XDK Readers: Creating the X-Men, How It All Beg...Michael Teitelbaum2000DK Publishing Inc
2216780789466953DK Readers: Creating the X-Men, How Comic Book...James Buckley2000DK Publishing Inc
## 继续纠正出版年鉴的类型
books.yearOfPublication=pd.to_numeric(books.yearOfPublication, errors='coerce')
sorted(books['yearOfPublication'].unique())

[0.0,
 1376.0,
 1378.0,
 1806.0,
 1897.0,
 1900.0,
 1901.0,
 1902.0,
 1904.0,
 1906.0,
 1908.0,
 1909.0,
 1910.0,
 1911.0,
 1914.0,
 1917.0,
 1919.0,
 1920.0,
 1921.0,
 1922.0,
 1923.0,
 1924.0,
 1925.0,
 1926.0,
 1927.0,
 1928.0,
 1929.0,
 1930.0,
 1931.0,
 1932.0,
 1933.0,
 1934.0,
 1935.0,
 1936.0,
 1937.0,
 1938.0,
 1939.0,
 1940.0,
 1941.0,
 1942.0,
 1943.0,
 1944.0,
 1945.0,
 1946.0,
 1947.0,
 1948.0,
 1949.0,
 1950.0,
 1951.0,
 1952.0,
 1953.0,
 1954.0,
 1955.0,
 1956.0,
 1957.0,
 1958.0,
 1959.0,
 1960.0,
 1961.0,
 1962.0,
 1963.0,
 1964.0,
 1965.0,
 1966.0,
 1967.0,
 1968.0,
 1969.0,
 1970.0,
 1971.0,
 1972.0,
 1973.0,
 1974.0,
 1975.0,
 1976.0,
 1977.0,
 1978.0,
 1979.0,
 1980.0,
 1981.0,
 1982.0,
 1983.0,
 1984.0,
 1985.0,
 1986.0,
 1987.0,
 1988.0,
 1989.0,
 1990.0,
 1991.0,
 1992.0,
 1993.0,
 1994.0,
 1995.0,
 1996.0,
 1997.0,
 1998.0,
 1999.0,
 2000.0,
 2001.0,
 2002.0,
 2003.0,
 2004.0,
 2005.0,
 2006.0,
 2008.0,
 2010.0,
 2011.0,
 2012.0,
 2020.0,
 2021.0,
 2024.0,
 2026.0,
 2030.0,
 2037.0,
 2038.0,
 2050.0,
 nan]

## 现在可以看出yearOfPublication的类型为int，其值范围为0-2050。

## 由于该数据集建于2004年，我假设2006年之后的所有年份都无效，保留两年的保证金，以防数据集可能已更新。

## 对于所有无效条目（包括0），我将这些条目转换为NaN，然后​​用剩余年份的平均值替换它们。
books.loc[(books.yearOfPublication > 2006) | (books.yearOfPublication == 0),'yearOfPublication'] = np.NAN

# 用年出版的平均价值代替NaNs在案例数据集被更新的情况下保留一定的空白
books.yearOfPublication.fillna(round(books.yearOfPublication.mean()), inplace=True)
books.yearOfPublication.isnull().sum()

0

books.yearOfPublication = books.yearOfPublication.astype(np.int32)

## publisher
books.loc[books.publisher.isnull(),:]
ISBNbookTitlebookAuthoryearOfPublicationpublisher
128890193169656XTyrant MoonElaine Corvidae2002NaN
1290371931696993Finders KeepersLinnea Sinclair2001NaN
## 检查行是否有书签作为查找器，看看我们是否能得到任何线索

## 与不同的出版商和图书作者的所有行
books.loc[(books.bookTitle == 'Tyrant Moon'),:]
ISBNbookTitlebookAuthoryearOfPublicationpublisher
128890193169656XTyrant MoonElaine Corvidae2002NaN
books.loc[(books.bookTitle == 'Finders Keepers'),:]
ISBNbookTitlebookAuthoryearOfPublicationpublisher
10799082177364XFinders KeepersFern Michaels2002Zebra Books
420190070465037Finders KeepersBarbara Nickolae1989McGraw-Hill Companies
582640688118461Finders KeepersEmily Rodda1993Harpercollins Juvenile Books
666781575663236Finders KeepersFern Michaels1998Kensington Publishing Corporation
1290371931696993Finders KeepersLinnea Sinclair2001NaN
1343090156309505Finders KeepersWill1989Voyager Books
1734730973146907Finders KeepersSean M. Costello2002Red Tower Publications
1958850061083909Finders KeepersSharon Sala2003HarperTorch
2118740373261160Finders KeepersElizabeth Travis1993Worldwide Library
## 由图书作者检查以找到模式

## 都有不同的出版商。这里没有线索
books.loc[(books.bookAuthor == 'Elaine Corvidae'),:]
ISBNbookTitlebookAuthoryearOfPublicationpublisher
1267621931696934Winter's OrphansElaine Corvidae2001Novelbooks
128890193169656XTyrant MoonElaine Corvidae2002NaN
1290010759901880WolfkinElaine Corvidae2001Hard Shell Word Factory
## 由图书作者检查以找到模式
books.loc[(books.bookAuthor == 'Linnea Sinclair'),:]
ISBNbookTitlebookAuthoryearOfPublicationpublisher
1290371931696993Finders KeepersLinnea Sinclair2001NaN
## 因为没有什么共同的东西可以推断出NaNs的发布者，将它们替换为“other”
books.loc[(books.ISBN == '193169656X'),'publisher'] = 'other'
books.loc[(books.ISBN == '1931696993'),'publisher'] = 'other'

## 二、用户数据集
print (users.shape)
users.head()

(278858, 3)
userIDLocationAge
01nyc, new york, usaNaN
12stockton, california, usa18.0
23moscow, yukon territory, russiaNaN
34porto, v.n.gaia, portugal17.0
45farnborough, hants, united kingdomNaN
users.dtypes

userID        int64
Location     object
Age         float64
dtype: object

users.userID.values

array([     1,      2,      3, ..., 278856, 278857, 278858], dtype=int64)

## Age 
sorted(users.Age.unique())

[nan,
 0.0,
 1.0,
 2.0,
 3.0,
 4.0,
 5.0,
 6.0,
 7.0,
 8.0,
 9.0,
 10.0,
 11.0,
 12.0,
 13.0,
 14.0,
 15.0,
 16.0,
 17.0,
 18.0,
 19.0,
 20.0,
 21.0,
 22.0,
 23.0,
 24.0,
 25.0,
 26.0,
 27.0,
 28.0,
 29.0,
 30.0,
 31.0,
 32.0,
 33.0,
 34.0,
 35.0,
 36.0,
 37.0,
 38.0,
 39.0,
 40.0,
 41.0,
 42.0,
 43.0,
 44.0,
 45.0,
 46.0,
 47.0,
 48.0,
 49.0,
 50.0,
 51.0,
 52.0,
 53.0,
 54.0,
 55.0,
 56.0,
 57.0,
 58.0,
 59.0,
 60.0,
 61.0,
 62.0,
 63.0,
 64.0,
 65.0,
 66.0,
 67.0,
 68.0,
 69.0,
 70.0,
 71.0,
 72.0,
 73.0,
 74.0,
 75.0,
 76.0,
 77.0,
 78.0,
 79.0,
 80.0,
 81.0,
 82.0,
 83.0,
 84.0,
 85.0,
 86.0,
 87.0,
 88.0,
 89.0,
 90.0,
 91.0,
 92.0,
 93.0,
 94.0,
 95.0,
 96.0,
 97.0,
 98.0,
 99.0,
 100.0,
 101.0,
 102.0,
 103.0,
 104.0,
 105.0,
 106.0,
 107.0,
 108.0,
 109.0,
 110.0,
 111.0,
 113.0,
 114.0,
 115.0,
 116.0,
 118.0,
 119.0,
 123.0,
 124.0,
 127.0,
 128.0,
 132.0,
 133.0,
 136.0,
 137.0,
 138.0,
 140.0,
 141.0,
 143.0,
 146.0,
 147.0,
 148.0,
 151.0,
 152.0,
 156.0,
 157.0,
 159.0,
 162.0,
 168.0,
 172.0,
 175.0,
 183.0,
 186.0,
 189.0,
 199.0,
 200.0,
 201.0,
 204.0,
 207.0,
 208.0,
 209.0,
 210.0,
 212.0,
 219.0,
 220.0,
 223.0,
 226.0,
 228.0,
 229.0,
 230.0,
 231.0,
 237.0,
 239.0,
 244.0]

## 年龄栏有一些无效的条目，比如nan，0和非常高的值，比如100和以上
users.loc[(users.Age > 90) | (users.Age < 5), 'Age'] = np.nan

## 用平均值代替NaN
## 将数据类型设置为int
users.Age = users.Age.fillna(users.Age.mean())
users.Age = users.Age.astype(np.int32)

sorted(users.Age.unique())

[5,
 6,
 7,
 8,
 9,
 10,
 11,
 12,
 13,
 14,
 15,
 16,
 17,
 18,
 19,
 20,
 21,
 22,
 23,
 24,
 25,
 26,
 27,
 28,
 29,
 30,
 31,
 32,
 33,
 34,
 35,
 36,
 37,
 38,
 39,
 40,
 41,
 42,
 43,
 44,
 45,
 46,
 47,
 48,
 49,
 50,
 51,
 52,
 53,
 54,
 55,
 56,
 57,
 58,
 59,
 60,
 61,
 62,
 63,
 64,
 65,
 66,
 67,
 68,
 69,
 70,
 71,
 72,
 73,
 74,
 75,
 76,
 77,
 78,
 79,
 80,
 81,
 82,
 83,
 84,
 85,
 86,
 87,
 88,
 89,
 90]

## 三、评级数据集
ratings.shape

(1149780, 3)

## 如果每个用户对每个条目进行评级，那么评级数据集将有nusers * nbooks条目，这表明数据集非常稀疏。
n_users = users.shape[0]
n_books = books.shape[0]
print (n_users * n_books)

75670906880

ratings.head(5)
userIDISBNbookRating
0276725034545104X0
127672601550612245
227672704465208020
3276729052165615X3
427672905217950286
ratings.bookRating.unique()

array([ 0,  5,  3,  6,  8,  7, 10,  9,  4,  1,  2], dtype=int64)

ratings_new = ratings[ratings.ISBN.isin(books.ISBN)]

print (ratings.shape)
print (ratings_new.shape)

(1149780, 3)
(1031136, 3)

## 没有新用户添加，因此我们将使用高于数据集的新用户（1031136，3）
print ("number of users: " + str(n_users))
print ("number of books: " + str(n_books))

number of users: 278858
number of books: 271360

sparsity=1.0-len(ratings_new)/float(n_users*n_books)
print ('图书交叉数据集的稀疏级别是 ' +  str(sparsity*100) + ' %')

图书交叉数据集的稀疏级别是 99.99863734155898 %

ratings.bookRating.unique()

array([ 0,  5,  3,  6,  8,  7, 10,  9,  4,  1,  2], dtype=int64)

ratings_explicit = ratings_new[ratings_new.bookRating != 0]
ratings_implicit = ratings_new[ratings_new.bookRating == 0]

print (ratings_new.shape)
print( ratings_explicit.shape)
print (ratings_implicit.shape)

(1031136, 3)
(383842, 3)
(647294, 3)

## 统计
sns.countplot(data=ratings_explicit , x='bookRating')
plt.show()

在这里插入图片描述

## 基于简单流行度的推荐系统
ratings_count = pd.DataFrame(ratings_explicit.groupby(['ISBN'])['bookRating'].sum())
top10 = ratings_count.sort_values('bookRating', ascending = False).head(10)
print ("推荐下列书籍")
top10.merge(books, left_index = True, right_on = 'ISBN')

推荐下列书籍
bookRatingISBNbookTitlebookAuthoryearOfPublicationpublisher
40857870316666343The Lovely Bones: A NovelAlice Sebold2002Little, Brown
74841080385504209The Da Vinci CodeDan Brown2003Doubleday
52231340312195516The Red Tent (Bestselling Backlist)Anita Diamant1998Picador USA
21432798059035342XHarry Potter and the Sorcerer's Stone (Harry P...J. K. Rowling1999Arthur A. Levine Books
35625950142001740The Secret Life of BeesSue Monk Kidd2003Penguin Books
2625510971880107Wild AnimusRich Shapero2004Too Far
110525240060928336Divine Secrets of the Ya-Ya Sisterhood: A NovelRebecca Wells1997Perennial
70624020446672211Where the Heart Is (Oprah's Book Club (Paperba...Billie Letts1998Warner Books
23122190452282152Girl with a Pearl EarringTracy Chevalier2001Plume Books
11821790671027360Angels &amp; DemonsDan Brown2001Pocket Star
users_exp_ratings = users[users.userID.isin(ratings_explicit.userID)]
users_imp_ratings = users[users.userID.isin(ratings_implicit.userID)]

print (users.shape)
print (users_exp_ratings.shape)
print (users_imp_ratings.shape)

(278858, 3)
(68091, 3)
(52451, 3)

## 基于协同过滤的推荐系统
counts1 = ratings_explicit['userID'].value_counts()
ratings_explicit = ratings_explicit[ratings_explicit['userID'].isin(counts1[counts1 >= 100].index)]
counts = ratings_explicit['bookRating'].value_counts()
ratings_explicit = ratings_explicit[ratings_explicit['bookRating'].isin(counts[counts >= 100].index)]

ratings_matrix = ratings_explicit.pivot(index='userID', columns='ISBN', values='bookRating')
userID = ratings_matrix.index
ISBN = ratings_matrix.columns
print(ratings_matrix.shape)
ratings_matrix.head()

(449, 66574)
ISBN00009131540001046438000104687X00010472130001047973000104799X0001048082000105373600010537440001055607...B000092Q0AB00009EF82B00009NDANB0000DYXIDB0000T6KHIB0000VZEJQB0000X8HIEB00013AX9EB0001I1KOGB000234N3A
userID
2033NaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaN
2110NaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaN
2276NaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaN
4017NaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaN
4385NaNNaNNaNNaNNaNNaNNaNNaNNaNNaN...NaNNaNNaNNaNNaNNaNNaNNaNNaNNaN

5 rows × 66574 columns

n_users = ratings_matrix.shape[0] #只考虑那些给出明确评级的用户
n_books = ratings_matrix.shape[1]
print (n_users, n_books)

449 66574

ratings_matrix.fillna(0, inplace = True)
ratings_matrix = ratings_matrix.astype(np.int32)

ratings_matrix.head(5)
ISBN00009131540001046438000104687X00010472130001047973000104799X0001048082000105373600010537440001055607...B000092Q0AB00009EF82B00009NDANB0000DYXIDB0000T6KHIB0000VZEJQB0000X8HIEB00013AX9EB0001I1KOGB000234N3A
userID
20330000000000...0000000000
21100000000000...0000000000
22760000000000...0000000000
40170000000000...0000000000
43850000000000...0000000000

5 rows × 66574 columns

sparsity=1.0-len(ratings_explicit)/float(users_exp_ratings.shape[0]*n_books)
print ('图书交叉数据集的稀疏级别是 ' +  str(sparsity*100) + ' %')

图书交叉数据集的稀疏级别是 99.99772184106935 %

## 基于用户的协同过滤
global metric,k
k=10
metric='cosine'
def findksimilarusers(user_id, ratings, metric = metric, k=k):
    similarities=[]
    indices=[]
    model_knn = NearestNeighbors(metric = metric, algorithm = 'brute') 
    model_knn.fit(ratings)
    loc = ratings.index.get_loc(user_id)
    distances, indices = model_knn.kneighbors(ratings.iloc[loc, :].values.reshape(1, -1), n_neighbors = k+1)
    similarities = 1-distances.flatten()
            
    return similarities,indices

def predict_userbased(user_id, item_id, ratings, metric = metric, k=k):
    prediction=0
    user_loc = ratings.index.get_loc(user_id)
    item_loc = ratings.columns.get_loc(item_id)
    similarities, indices=findksimilarusers(user_id, ratings,metric, k) #similar users based on cosine similarity
    mean_rating = ratings.iloc[user_loc,:].mean() #to adjust for zero based indexing
    sum_wt = np.sum(similarities)-1
    product=1
    wtd_sum = 0 
    
    for i in range(0, len(indices.flatten())):
        if indices.flatten()[i] == user_loc:
            continue;
        else: 
            ratings_diff = ratings.iloc[indices.flatten()[i],item_loc]-np.mean(ratings.iloc[indices.flatten()[i],:])
            product = ratings_diff * (similarities[i])
            wtd_sum = wtd_sum + product
    
    #在非常稀疏的数据集的情况下，使用基于协作的方法的相关度量可能会给出负面的评价
    #在这里的处理如下
    if prediction <= 0:
        prediction = 1   
    elif prediction >10:
        prediction = 10
    
    prediction = int(round(mean_rating + (wtd_sum/sum_wt)))
    print ('用户预测等级 {0} -> item {1}: {2}'.format(user_id,item_id,prediction))
 
    return prediction

## 测试
predict_userbased(11676,'0001056107',ratings_matrix)

用户预测等级 11676 -> item 0001056107: 2





2

## 基于项目的协同过滤
def findksimilaritems(item_id, ratings, metric=metric, k=k):
    similarities=[]
    indices=[]
    ratings=ratings.T
    loc = ratings.index.get_loc(item_id)
    model_knn = NearestNeighbors(metric = metric, algorithm = 'brute')
    model_knn.fit(ratings)
    
    distances, indices = model_knn.kneighbors(ratings.iloc[loc, :].values.reshape(1, -1), n_neighbors = k+1)
    similarities = 1-distances.flatten()
 
    return similarities,indices

def predict_itembased(user_id, item_id, ratings, metric = metric, k=k):
    prediction= wtd_sum =0
    user_loc = ratings.index.get_loc(user_id)
    item_loc = ratings.columns.get_loc(item_id)
    similarities, indices=findksimilaritems(item_id, ratings) #similar users based on correlation coefficients
    sum_wt = np.sum(similarities)-1
    product=1
    for i in range(0, len(indices.flatten())):
        if indices.flatten()[i] == item_loc:
            continue;
        else:
            product = ratings.iloc[user_loc,indices.flatten()[i]] * (similarities[i])
            wtd_sum = wtd_sum + product                              
    prediction = int(round(wtd_sum/sum_wt))
    
    #在非常稀疏的数据集的情况下，使用基于协作的方法的相关度量可能会给出负面的评价
    #在这里处理的是下面的//代码，没有下面的代码片段，下面的代码片段是为了避免负面影响
    #在使用相关度规时，可能会出现非常稀疏的数据集的预测
    if prediction <= 0:
        prediction = 1   
    elif prediction >10:
        prediction = 10
 
    print ('用户预测等级 {0} -> item {1}: {2}'.format(user_id,item_id,prediction)    )  
    
    return prediction

## 测试
prediction = predict_itembased(11676,'0001056107',ratings_matrix)

用户预测等级 11676 -> item 0001056107: 1

本文来自互联网用户投稿，该文观点仅代表作者本人，不代表本站立场。本站仅提供信息存储空间服务，不拥有所有权，不承担相关法律责任。如若转载，请注明出处：http://www.coloradmin.cn/o/939566.html

如若内容造成侵权/违法违规/事实不符，请联系多彩编程网进行投诉反馈，一经查实，立即删除！

相关文章

【Python|第40期】Python操作MongoDB

【Python|第40期】Python操作MongoDB

日期&#xff1a;2023年8月28日 作者&#xff1a;Commas 签名&#xff1a;(ง •_•)ง 积跬步以致千里,积小流以成江海…… 注释&#xff1a;如果您觉得有所帮助&#xff0c;帮忙点个赞&#xff0c;也可以关注我&#xff0c;我们一起成长&#xff1b;如果有不对的地方&#xf…
阅读更多...
一步一步实验，讲解python中模块和包的使用

一步一步实验，讲解python中模块和包的使用

背景 为什么要提出这个问题&#xff1f; 在一个项目中&#xff0c;每一个python文件打开后&#xff0c;都会看到依赖了其他的一些包、模块等&#xff1b;概念混乱&#xff0c;魔改目标不清晰 为什么要修改&#xff1f; 如果需要将某开源包进行自定义处理&#xff0c;不再使…
阅读更多...
AI智慧，书香飘溢|在图书馆的世界，知识与未来交汇！

AI智慧，书香飘溢|在图书馆的世界，知识与未来交汇！

“如果有天堂&#xff0c;天堂应该是图书馆的模样。”阿根廷国家图书馆前馆长、著名作家博尔赫斯如此形容图书馆。人类文明的演进与传承&#xff0c;倚仗于知识的积累&#xff0c;而知识的载体往往绕不开书籍&#xff0c;其集散地——图书馆&#xff0c;更是在这一过程扮演着极…
阅读更多...
【精算研究01/10】 计量经济学的性质和范围

【精算研究01/10】 计量经济学的性质和范围

一、说明 计量经济学是使用统计方法来发展理论或测试经济学或金融学中的现有假设。计量经济学依赖于回归模型和零假设检验等技术。计量经济学也可以用来预测未来的经济或金融趋势。 图片来源&#xff1a;https://marketbusinessnews.com 二、 计量经济之简介 计量经济学是对经济…
阅读更多...
javacv基础02-调用本机摄像头并预览摄像头图像画面视频

javacv基础02-调用本机摄像头并预览摄像头图像画面视频

引入架包&#xff1a; <dependency><groupId>org.openpnp</groupId><artifactId>opencv</artifactId><version>4.5.5-1</version></dependency><dependency><groupId>org.bytedeco</groupId><artifactId…
阅读更多...
【校招VIP】算法考点之堆排

【校招VIP】算法考点之堆排

考点介绍&#xff1a; 排序算法属于数据结构和算法的基础内容&#xff0c;并且也是大厂笔试中的高频考点。 堆排序是使用一棵树存储序列这个课树只保证跟节点是这棵树中的最小值&#xff0c;但并不保证其他节点是按顺序的。因此他的排序是每次从堆中取得堆顶&#xff0c;取得 n…
阅读更多...
Qt Scroll Area控件设置，解决无法显示全部内容，且无法滚动显示问题。

Qt Scroll Area控件设置，解决无法显示全部内容，且无法滚动显示问题。

前言&#xff0c;因为要显示很多条目的内容&#xff0c;原来是用Vertical Layout控件里面嵌套Horizontal layout显示了很多行控件&#xff0c;发现最简单的方法就是使用滚动条控件&#xff0c;但是无论如何调整需要滚动的控件高度&#xff0c;始终无法滚动显示内容。也就是说添…
阅读更多...
5. 线性层及其他层

5. 线性层及其他层

5.1 神经网络结构 5.2 线性拉平 import torch import torchvision from torch import nn from torch.nn import ReLU from torch.nn import Sigmoid from torch.utils.data import DataLoader from torch.utils.tensorboard import SummaryWriterdataset torchvision.datase…
阅读更多...
ant-vue1.78版监听a-modal遮罩层的滚动事件

ant-vue1.78版监听a-modal遮罩层的滚动事件

监听a-modal遮罩层的滚动事件 我们开发过程中经常有遇到监听页面滚动的事件需求&#xff0c;去做一些下拉加载或者是下拉分页的需求&#xff0c;我们直接在vue的生命周期中去绑定事件监听非常的方便&#xff0c;但如果是弹框的遮罩层的滚动监听呢&#xff1f;页面的监听完全是…
阅读更多...
Stone Prover：StarkWare的STARK Prover

Stone Prover：StarkWare的STARK Prover

1. 引言 StarkWare将基于Apache 2.0 license&#xff0c;开源其以C编写的STARK Prover&#xff0c;名为Stone&#xff08;STARK one&#xff09;。 其基本流程为&#xff1a; 1&#xff09;编写Cairo0程序。2&#xff09;使用Cairo工具 将Cairo0程序编译为CASM。3&#xff…
阅读更多...
数据结构：八种数据结构大全

数据结构：八种数据结构大全

数据结构 1.1 数据结构概述 数据结构是计算机存储、组织数据的方式&#xff1b;通常情况下&#xff0c;精心选择的数据结构可以带来更高的运行或者存储效率。数据结构的优良将直接影响着我们程序的性能&#xff1b;常用的数据结构有&#xff1a;数组&#xff08;Array&#xff…
阅读更多...
K8S Nginx Ingress实现金丝雀发布

K8S Nginx Ingress实现金丝雀发布

通过给 Ingress 资源指定 Nginx Ingress 所支持的 annotation 可实现金丝雀发布。 需给服务创建2个 Ingress&#xff0c;其中1个常规 Ingress&#xff0c;另1个为带 nginx.ingress.kubernetes.io/canary: "true" 固定的 annotation 的 Ingress&#xff0c;称为 Cana…
阅读更多...
浅谈Lua协程和函数的尾调用

浅谈Lua协程和函数的尾调用

前言 虽然不经常用到协程&#xff0c;但是也不能谈虎色变。同时&#xff0c;在有些场景&#xff0c;协程会起到一种不可比拟的作用。所以&#xff0c;了解它&#xff0c;对于一些功能&#xff0c;也会有独特的思路和想法。 协程 概念 关于进程和线程的概念就不多说。 那么…
阅读更多...
这所985专业课均分130！复试笔试很难！淘汰率很高！

这所985专业课均分130！复试笔试很难！淘汰率很高！

一、学校及专业介绍 东南大学坐落于六朝古都南京&#xff0c;是享誉海内外的著名高等学府。学校是国家教育部直属并与江苏省共建的全国重点大学&#xff0c;是国家“双一流”、“985工程”、“211工程”重点建设高校。2017年&#xff0c;东南大学入选世界一流大学建设名单。 …
阅读更多...
字符集（Latin1，GBK，utf8，utf8mb4）

字符集（Latin1，GBK，utf8，utf8mb4）

Latin1 1个字符占一个字节GBK 1个字符占两个字节utf8utfmb3 1个字节占三个字节utf8mb4 1个字符占四个字节
阅读更多...
OLED透明屏高清：什么是OLED透明屏？未来显示技术的巅峰之作

OLED透明屏高清：什么是OLED透明屏？未来显示技术的巅峰之作

在现代科技快速发展的时代&#xff0c;高清显示已经成为人们对于视觉体验的基本要求。 而OLED透明屏作为一种先进的显示技术&#xff0c;以其出色的高清显示效果和透明度&#xff0c;正逐渐成为未来显示技术的巅峰之作。 一、什么是OLED透明屏 OLED透明屏采用有机发光二极管技…
阅读更多...
Zabbix下载安装及SNMP Get使用

Zabbix下载安装及SNMP Get使用

帮助文档&#xff1a;6. Zabbix Appliance 一、zabbix下载安装 1、获取Zabbix Appliance镜像 Download Zabbix appliance 2、使用该镜像创建虚拟机 3、打开虚拟机控制台自动安装&#xff0c;等待安装完成即可 默认配置 系统/数据库&#xff1a;root:zabbix Zabbix 前端&am…
阅读更多...
javacv基础01-HelloWorld

javacv基础01-HelloWorld

JavaCV是一个针对Java编程语言的开源计算机视觉和机器学习库。它为各种流行的计算机视觉和图像处理库&#xff08;如OpenCV、FFmpeg等&#xff09;提供了Java包装&#xff0c;使Java开发人员能够在其Java应用程序中访问这些库的功能&#xff0c;无需编写本地代码。 JavaCV提供…
阅读更多...
SpringBoot Mybatis 多数据源 MySQL+Oracle

SpringBoot Mybatis 多数据源 MySQL+Oracle

一、背景 在SpringBoot Mybatis 项目中&#xff0c;需要连接 多个数据源&#xff0c;连接多个数据库&#xff0c;需要连接一个MySQL数据库和一个Oracle数据库 二、依赖 pom.xml <dependencies><dependency><groupId>org.springframework.boot</groupId&…
阅读更多...
（Deep Learning）准确率和召回率的基础概念

（Deep Learning）准确率和召回率的基础概念

算法模型极大的提升了对各类结果的预测效率。 【算法模型的本质】 算法模型的本质&#xff0c;是基于输入的各类变量因子&#xff0c;通过计算规则&#xff08;模型or公式&#xff09;&#xff0c;得出预测结果。 典型的预测结果比如&#xff1a; 1.&#xff08;通过历史行为…
阅读更多...
推荐文章
最新文章

Copyright @ 2022~2023