RMSE MAE pre
q=10 k=20 p对MAE的影响 100K
[25, 50, 75, 100, 125, 150] [0.9532408116247819, 0.9483862493801819, 0.9465119037173325, 0.9462897732497364, 0.946237384787047, 0.9466131897479011]
[25, 50, 75, 100, 125, 150] [0.750336692756152, 0.7466744068502496, 0.7449879941045279, 0.7445761522147301, 0.7443785467682291, 0.7446004315981095]
[25, 50, 75, 100, 125, 150] [0.18449628844114527, 0.1690986214209968, 0.16182396606574762, 0.16019088016967126, 0.1591728525980912, 0.16031813361611877]
q=10 k=20 p对MAE的影响 1M
[25, 50, 75, 100, 125, 150] [0.930184887209983, 0.9247703864553927, 0.922592364423042, 0.9214491808568143, 0.9207574479986459, 0.9204712471665468]
[25, 50, 75, 100, 125, 150] [0.7308832082588673, 0.7268606959624528, 0.7251939587376949, 0.7243831300040574, 0.7238817123879806, 0.7236672716684257]
[25, 50, 75, 100, 125, 150] [0.18756622516556293, 0.17955298013245033, 0.1762019867549669, 0.17350000000000002, 0.17329470198675498, 0.17356622516556292]
p=125 q=20 k对MAE的影响 100K
[5, 10, 15, 20, 25, 30] [0.9703415399184427, 0.9522638693527069, 0.9474344859733878, 0.9460303561062051, 0.9457864612874911, 0.9454072003049937]
[5, 10, 15, 20, 25, 30] [0.7628872220973382, 0.7488155302673848, 0.7450007296865588, 0.7442517044332744, 0.744222110626528, 0.7439334623349032]
[5, 10, 15, 20, 25, 30] [0.13713679745493107, 0.15283138918345704, 0.15700954400848355, 0.15889713679745493, 0.15779427359490986, 0.1586426299045599]
p=125 q=20 k对MAE的影响 1M
[5, 10, 15, 20, 25, 30] [0.9470681430189858, 0.9286533831803702, 0.922994755173866, 0.9206025414396919, 0.9194386476050758, 0.9186616154766188]
[5, 10, 15, 20, 25, 30] [0.7445727875950261, 0.7299466545411931, 0.7256479642923972, 0.7237744732922933, 0.7229525940732703, 0.7223160211233509]
[5, 10, 15, 20, 25, 30] [0.13876490066225167, 0.159341059602649, 0.1681390728476821, 0.1729437086092715, 0.17573841059602652, 0.1776456953642384]
p=125 k=30 q对MAE的影响 100k
q要x5
[5, 10, 15, 20, 25, 30] [0.9453495919386089, 0.9449879165517545, 0.9446974086316292, 0.944487782625945, 0.9442526953592795, 0.9440948152244545]
[5, 10, 15, 20, 25, 30] [0.7439082401200476, 0.7437017329025323, 0.7435446911105427, 0.7434274543664485, 0.7433098208516034, 0.7432191862877311]
[5, 10, 15, 20, 25, 30] [0.15849416755037118, 0.15800636267232238, 0.1575609756097561, 0.1571367974549311, 0.15692470837751854, 0.15671261930010605]
[5, 10, 15, 20, 25, 30] [0.9457170669670802, 0.9455998791938622, 0.9454930317753651, 0.9454072003049937, 0.9453495919386089, 0.9452628789681894]
[5, 10, 15, 20, 25, 30] [0.7441156930918288, 0.7440432755470274, 0.7439860548712594, 0.7439334623349032, 0.7439082401200476, 0.7438632943080367]
[5, 10, 15, 20, 25, 30] [0.1591728525980912, 0.15889713679745493, 0.15881230116648992, 0.1586426299045599, 0.15849416755037118, 0.15843054082714741]
p=125 k=30 q对MAE的影响 1M
q要x5
[5, 10, 15, 20, 25, 30] [0.9185882389376359, 0.9182077106100592, 0.9178778536897214, 0.9175575626635741, 0.9172465218271272, 0.9169489822035526]
[5, 10, 15, 20, 25, 30] [0.7222653953419933, 0.7220211843081594, 0.7218238409472799, 0.7216299533580426, 0.7214385319082218, 0.7212578104938808]
[5, 10, 15, 20, 25, 30] [0.17746026490066225, 0.17674172185430465, 0.1762185430463576, 0.1757682119205298, 0.17539403973509932, 0.17505298013245033]
[5, 10, 15, 20, 25, 30] [0.9188958646583396, 0.918815651095026, 0.9187401064297147, 0.9186614479540992, 0.9185880793515964, 0.9185167091028605]
[5, 10, 15, 20, 25, 30] [0.7224719434913467, 0.7224183672338095, 0.7223696860035796, 0.7223157935514628, 0.7222651772450416, 0.722220112825609]
[5, 10, 15, 20, 25, 30] [0.17827152317880796, 0.17811920529801323, 0.17787417218543045, 0.17764238410596025, 0.17745695364238415, 0.17727152317880795]
100k对比试验:
k=[5, 10, 15, 20, 25, 30]
对比四个
hfcf
p=125 q=30
[5, 10, 15, 20, 25, 30] [0.9702417174397, 0.952150654001618, 0.9472765275972442, 0.9458851971401968, 0.9456616042265269, 0.9452599130819592]
[5, 10, 15, 20, 25, 30] [0.7628071916104936, 0.74877614027437, 0.7449177608601305, 0.744154815133356, 0.744160436924379, 0.7438601928000093]
[5, 10, 15, 20, 25, 30] [0.13671261930010603, 0.15261930010604458, 0.15660657476139977, 0.1584729586426299, 0.15745493107104985, 0.15843054082714741]
p=125 q=40
[5, 10, 15, 20, 25, 30] [0.9702768911543451, 0.9520272076173122, 0.9471639741291267, 0.9457130966089491, 0.9454828449116084, 0.9451235183537428]
[5, 10, 15, 20, 25, 30] [0.7628478727645928, 0.7486843429719707, 0.7448341282350835, 0.744064892139911, 0.7440634129427722, 0.7437784650198106]
[5, 10, 15, 20, 25, 30] [0.13633085896076352, 0.15236479321314952, 0.15633085896076354, 0.15836691410392362, 0.15720042417815483, 0.15819724284199363]
p=125 q=80
[5, 10, 15, 20, 25, 30] [0.9699229369807363, 0.9516354057321038, 0.9466986602970081, 0.9452978331891988, 0.9450538037073389, 0.9446483083184212]
[5, 10, 15, 20, 25, 30] [0.7626535868917514, 0.7484760004542632, 0.7445541942958223, 0.7438292410435712, 0.7438310077928902, 0.7435162116748997]
[5, 10, 15, 20, 25, 30] [0.13484623541887591, 0.15158006362672322, 0.1554612937433722, 0.15758218451749734, 0.15639448568398728, 0.1574337221633086]
p=125 q=150
[5, 10, 15, 20, 25, 30] [0.969669707859613, 0.9511272253927018, 0.9461432989177535, 0.944789691991803, 0.9444967670976432, 0.9440919918992974]
[5, 10, 15, 20, 25, 30] [0.7625454670211639, 0.7482147229201692, 0.7441599229301129, 0.7435519435274802, 0.7435350546513638, 0.7432164079020238]
[5, 10, 15, 20, 25, 30] [0.13219512195121952, 0.15022269353128315, 0.15437963944856842, 0.1567338282078473, 0.15584305408271476, 0.15671261930010605]
ub
[5, 10, 15, 20, 25, 30] [1.1581161841979928, 1.1153105459173829, 1.0872370349396527, 1.0716079616846625, 1.0606364542210247, 1.05203871044538]
[5, 10, 15, 20, 25, 30] [0.8942434240828245, 0.8694929684305333, 0.8516889623114716, 0.8417937338158199, 0.8350480843419901, 0.8292706780746724]
[5, 10, 15, 20, 25, 30] [0.08704135737009544, 0.04407211028632026, 0.02740190880169671, 0.018685047720042418, 0.014231177094379637, 0.010286320254506893]
mean
[5, 10, 15, 20, 25, 30] [1.3528964179278886, 1.2508866623932369, 1.186913009051805, 1.1525559929694738, 1.1294325779751946, 1.1112601456997038]
[5, 10, 15, 20, 25, 30] [0.9040765770729962, 0.8749109869631475, 0.8552950526819988, 0.8449397580748311, 0.8380221235830936, 0.8320858440817464]
[5, 10, 15, 20, 25, 30] [0.09505832449628844, 0.046341463414634146, 0.02814422057264051, 0.01891834570519618, 0.013955461293743374, 0.009968186638388122]
mode
[5, 10, 15, 20, 25, 30] [1.5990267085847498, 1.5725668889892193, 1.5349809275715898, 1.5007143273515773, 1.4786524931540757, 1.4570692527380447]
[5, 10, 15, 20, 25, 30] [0.9520578384249381, 0.9327977247641105, 0.9151144471352444, 0.9021648513885576, 0.891713736667197, 0.8837036965026515]
[5, 10, 15, 20, 25, 30] [0.14016967126192997, 0.11580063626723222, 0.10644750795334042, 0.09832449628844114, 0.09374337221633086, 0.0896712619300106]
lfm
[5, 10, 15, 20, 25, 30] [0.9470183650608288, 0.9465495847030251, 0.944021097132195, 0.944088734329976, 0.9430330759010441, 0.9430534659265681]
[5, 10, 15, 20, 25, 30] [0.7465005323519327, 0.7464169721991285, 0.7441061249600162, 0.7445102476422552, 0.7434519111327264, 0.743358858978395]
[5, 10, 15, 20, 25, 30] [0.06356309650053024, 0.06258748674443267, 0.06663838812301166, 0.06381760339342524, 0.06606574761399787, 0.06506892895015906]
1M对比试验:
k=[5, 10, 15, 20, 25, 30]
对比四个
hfcf
p=125 q=150
[5, 10, 15, 20, 25, 30] [0.9459841446667749, 0.9271576584813281, 0.9213867465846313, 0.9189574590165511, 0.9177496080220852, 0.9169489822035526]
[5, 10, 15, 20, 25, 30] [0.7439305409688532, 0.7290594948635076, 0.7246675746630282, 0.7227701100462587, 0.7219196891097803, 0.7212578104938808]
[5, 10, 15, 20, 25, 30] [0.13473841059602648, 0.15607615894039734, 0.16512251655629137, 0.17026158940397348, 0.17320860927152318, 0.17505298013245033]
ub
[5, 10, 15, 20, 25, 30] [1.1273288654812657, 1.0811186055288307, 1.053482658196914, 1.035
没有合适的资源?快使用搜索试试~ 我知道了~
温馨提示
纯Python实现的协同过滤推荐系统(可作毕设).zip 【项目说明】 一个纯Python实现的协同过滤推荐系统。它包含了基于用户的协同过滤(UserCF)和基于物品的协同过滤(ItemCF)。同时还包括了随机推荐和最热门推荐作为对比。此外,还实现了著名的潜在因子模型(LFM)。 Main Function Points 创建训练集和测试集 训练推荐模型 给出推荐结果 评估推荐结果 Technology Stack 纯Python实现,不需要任何Python扩展库(如Numpy/Pandas)
资源推荐
资源详情
资源评论
收起资源包目录
纯Python实现的协同过滤推荐系统(可作毕设).zip (22个子文件)
utils.py 4KB
main.py 5KB
MeanValueCF.py 8KB
说明.7z 3KB
similarity.py 12KB
dataset.py 4KB
.idea
MovieLens-Recommender-master.iml 532B
other.xml 250B
vcs.xml 180B
misc.xml 189B
inspectionProfiles
Project_Default.xml 2KB
profiles_settings.xml 174B
modules.xml 435B
.gitignore 180B
ItemCF.py 8KB
random_pred.py 5KB
UserCF.py 8KB
HfCF.py 14KB
record.txt 9KB
ModeValueCF.py 8KB
LFM.py 8KB
most_popular.py 5KB
共 22 条
- 1
资源评论
.whl
- 粉丝: 3824
- 资源: 4664
上传资源 快速赚钱
- 我的内容管理 展开
- 我的资源 快来上传第一个资源
- 我的收益 登录查看自己的收益
- 我的积分 登录查看自己的积分
- 我的C币 登录后查看C币余额
- 我的收藏
- 我的下载
- 下载帮助
最新资源
资源上传下载、课程学习等过程中有任何疑问或建议,欢迎提出宝贵意见哦~我们会及时处理!
点击此处反馈
安全验证
文档复制为VIP权益,开通VIP直接复制
信息提交成功