1 00:00:00,000 --> 00:00:03,416 >> [Music kucheza] 2 00:00:03,416 --> 00:00:05,860 3 00:00:05,860 --> 00:00:08,180 >> Ubongo SCASSELLATI: Karibu kwa CS50 ai mfululizo. 4 00:00:08,180 --> 00:00:12,600 Jina langu ni Scass, na leo tunakwenda kuzungumzia mifumo MKUBALIAJI. 5 00:00:12,600 --> 00:00:15,780 Mifumo sasa MKUBALIAJI sauti kama aina ya jina isiyo ya kawaida. 6 00:00:15,780 --> 00:00:18,630 Inaonekana kama labda ni lazima kuwa mifumo mapendekezo, 7 00:00:18,630 --> 00:00:21,290 na mimi aina ya kukubaliana na wewe. 8 00:00:21,290 --> 00:00:26,110 Lakini haya ni mifumo ya kwamba msaada kuchagua nje mambo sawa wakati wowote 9 00:00:26,110 --> 00:00:28,210 kuchagua kitu online. 10 00:00:28,210 --> 00:00:32,119 Netflix, kwa mfano kupendekeza mengine sinema ambayo unaweza kutaka kuangalia. 11 00:00:32,119 --> 00:00:36,660 Au Pandora kupendekeza mbalimbali nyimbo ambazo unaweza kutaka kusikiliza. 12 00:00:36,660 --> 00:00:40,940 Amazon mapenzi zinaonyesha ni aina gani ya bidhaa nyingine unaweza kutaka kununua. 13 00:00:40,940 --> 00:00:43,520 Facebook hata kupendekeza baadhi ya marafiki wengine 14 00:00:43,520 --> 00:00:45,440 kwamba unaweza kutaka kuongeza. 15 00:00:45,440 --> 00:00:49,800 Kila moja ya mifumo hii hufanya kazi kwa kutumia sawa aina ya msingi ya algorithm, 16 00:00:49,800 --> 00:00:52,520 na hilo ndilo tuko kwenda kuzungumza kuhusu leo. 17 00:00:52,520 --> 00:00:56,860 >> Sasa algorithms hizi ni biashara ya kushangaza kubwa. 18 00:00:56,860 --> 00:01:01,130 Netflix miaka michache iliyopita katika 2009 inapatikana $ 1 milioni 19 00:01:01,130 --> 00:01:07,240 tuzo kama unaweza kuboresha zao mapendekezo ya mfumo na% 10 tu. 20 00:01:07,240 --> 00:01:11,960 Hiyo 10%, ingawa, inawakilisha kiasi kikubwa cha biashara. 21 00:01:11,960 --> 00:01:15,330 Makadirio ni vigumu kuja na, lakini watu wengi 22 00:01:15,330 --> 00:01:19,050 kuamini kwamba mapendekezo hayo mifumo ya ununuzi online 23 00:01:19,050 --> 00:01:25,729 mfumo kama Amazon kusababisha mahali fulani kati ya 10% na 25% ongezeko la mapato. 24 00:01:25,729 --> 00:01:27,770 Hivyo unaweza kufikiria aina ya kiasi kwamba wewe ni 25 00:01:27,770 --> 00:01:32,860 kuzungumza juu wakati sisi kufikiri juu hata hawa algorithms kidogo. 26 00:01:32,860 --> 00:01:35,200 >> Basi hebu kupata baadhi ya mifano. 27 00:01:35,200 --> 00:01:38,460 Jinsi ni kwamba hawa mifumo kweli kazi? 28 00:01:38,460 --> 00:01:40,773 Kuna mambo mawili ya msingi aina ya algorithms kwamba 29 00:01:40,773 --> 00:01:45,050 ni katika kucheza wakati sisi majadiliano juu kuzalisha mapendekezo. 30 00:01:45,050 --> 00:01:48,650 Ndio kwanza zinaitwa maudhui ya msingi ya kuchuja. 31 00:01:48,650 --> 00:01:53,410 Na maudhui ya msingi ya kuchuja hutegemea juu ya kufanana kati ya vitu 32 00:01:53,410 --> 00:02:00,370 wenyewe, kwamba ni kati ya sinema mbili au nyimbo mbili au vitu viwili kununuliwa. 33 00:02:00,370 --> 00:02:03,190 Tunakwenda kutumia sinema kama mfano, lakini hii 34 00:02:03,190 --> 00:02:07,850 inaweza kuomba, kwa kweli, kwa aina yoyote ya kupinga kwamba sisi ni kuangalia kwa. 35 00:02:07,850 --> 00:02:13,330 >> Hivyo kama nadhani kuhusu baadhi sinema kuanzia mwaka jana, 36 00:02:13,330 --> 00:02:16,799 Niliona Ndani Kati na watoto wangu, wao kupendwa hivyo. 37 00:02:16,799 --> 00:02:17,840 Lakini pia alikuwa na uchaguzi. 38 00:02:17,840 --> 00:02:21,350 Sisi inaweza kuwa wamekwenda kuona marafiki, tunaweza kuona Umri wa Ultron, 39 00:02:21,350 --> 00:02:24,850 au sisi inaweza kuwa na kuonekana Ant Man katika sinema. 40 00:02:24,850 --> 00:02:27,580 >> Kwa kimojawapo cha vitu hivi sinema, tunaweza kufikiria 41 00:02:27,580 --> 00:02:33,320 kuzalisha orodha ya makala au sifa kuhusu sinema wale mbalimbali. 42 00:02:33,320 --> 00:02:37,190 Hivyo kwa mfano, mimi naweza kufikiria yupi kati ya sinema wale ni animated. 43 00:02:37,190 --> 00:02:39,960 Naam, wote Kati Ndani na marafiki ni animated. 44 00:02:39,960 --> 00:02:44,140 Wala Umri wa Ultron wala Ant Man ni sinema animated. 45 00:02:44,140 --> 00:02:47,040 Na mimi naweza kufikiria jengo up muundo, meza kwamba 46 00:02:47,040 --> 00:02:49,440 unaorodhesha kila mmoja mali hizi. 47 00:02:49,440 --> 00:02:51,790 Je, wao ni animated au la? 48 00:02:51,790 --> 00:02:54,780 Mimi naweza kisha kuongeza zaidi makala ya meza hii 49 00:02:54,780 --> 00:02:58,380 kwa kuongeza safu zaidi katika muundo huu. 50 00:02:58,380 --> 00:03:00,970 Mimi naweza kuuliza kama au wao siyo sinema ajabu. 51 00:03:00,970 --> 00:03:04,010 Naam, Inside Out na marafiki si sinema ajabu, 52 00:03:04,010 --> 00:03:06,715 Umri wa Ultron na Ant Man hakika ni. 53 00:03:06,715 --> 00:03:09,100 >> Na mimi naweza kuuliza aina yoyote sifa ya tofauti 54 00:03:09,100 --> 00:03:12,080 kuwa nilitaka, aina yoyote ya makala hiyo inaweza kuwa ni muhimu kwangu. 55 00:03:12,080 --> 00:03:13,440 Je, wana villain super? 56 00:03:13,440 --> 00:03:16,700 Naam, hakuna villain super katika Ndani Nje, lakini kuna wale katika Marafiki 57 00:03:16,700 --> 00:03:19,990 na katika, ni wazi, mbili superhero sinema. 58 00:03:19,990 --> 00:03:23,900 >> Mimi naweza pia kuuliza mambo kama, vizuri, gani wao kupita mtihani Bechdel? 59 00:03:23,900 --> 00:03:27,280 Je, kuna wawili aitwaye wahusika wa kike ambao 60 00:03:27,280 --> 00:03:30,550 kutumia baadhi kiasi kikubwa cha wakati kuwa na mazungumzo kwamba 61 00:03:30,550 --> 00:03:34,400 haina kuhusisha wanaume katika kutupwa? 62 00:03:34,400 --> 00:03:39,870 Naam, katika kesi hii, Inside Out hupita mtihani, Marafiki inashindwa, Umri wa Ultron 63 00:03:39,870 --> 00:03:42,990 hupita mtihani, na Ant Man inashindwa. 64 00:03:42,990 --> 00:03:45,020 Mtu yeyote wa makala haya Mimi naweza kufikiria kuhusu 65 00:03:45,020 --> 00:03:48,660 kuwa ni muhimu kwa baadhi ya watu. 66 00:03:48,660 --> 00:03:52,000 >> Mimi naweza pia kuuliza mambo kama ni kuna watu wowote katika sinema hizi kwamba 67 00:03:52,000 --> 00:03:57,190 ni Mbegu kutoka hebu sema, Viwanja vya na Burudani, moja ya maonyesho favorite. 68 00:03:57,190 --> 00:04:00,540 Naam, Inside Out ina Amy Poehler, hiyo ni Alumni. 69 00:04:00,540 --> 00:04:01,530 Kwamba makosa. 70 00:04:01,530 --> 00:04:04,110 Jon Hamm alikuwa katika marafiki. 71 00:04:04,110 --> 00:04:08,600 Paul Rudd alikuwa katika Ant Man, lakini hakuna mtu katika Umri wa Ultron alikuwa katika Mbuga na Req 72 00:04:08,600 --> 00:04:10,150 vilevile. 73 00:04:10,150 --> 00:04:12,990 Hivyo siwezi kujenga orodha hii ya makala, lakini hawakuweza kweli 74 00:04:12,990 --> 00:04:14,710 kuwa kitu chochote kuhusu sinema. 75 00:04:14,710 --> 00:04:17,329 Wao wanaweza kuwa juu ya nini kipengele uwiano wao walipigwa risasi katika, 76 00:04:17,329 --> 00:04:21,630 inaweza kuwa jinsi viti wengi wao kuuzwa katika ufunguzi mwishoni mwa wiki zao. 77 00:04:21,630 --> 00:04:25,630 Kipengele yoyote ambayo nataka kuzalisha siwezi kuweka katika meza hii. 78 00:04:25,630 --> 00:04:29,600 >> Sasa, katika kesi hii, nimekuwa kujengwa kila aina ya maadili Bullion, 79 00:04:29,600 --> 00:04:33,700 ndiyo au hapana, kupita au kushindwa, lakini hawakuweza kuwa kitu chochote. 80 00:04:33,700 --> 00:04:36,690 Wao wanaweza kuwa maadili holela. 81 00:04:36,690 --> 00:04:39,070 Kwa maudhui ya msingi ya kuchuja, nini tunakwenda kufanya 82 00:04:39,070 --> 00:04:42,810 ni tunakwenda kufikiria nguzo mbili katika meza hii 83 00:04:42,810 --> 00:04:45,660 na kuona ni jinsi sawa walipo. 84 00:04:45,660 --> 00:04:48,640 Hivyo kwa mfano, kama mimi alikwenda kumuona Ndani Kati, 85 00:04:48,640 --> 00:04:53,640 Mimi kuuliza, ni nini sinema nyingine ili nipate kuwa tayari kwenda kuona. 86 00:04:53,640 --> 00:04:56,890 Hiyo ni, nini tayari kutumia fedha yangu kwenda kuona. 87 00:04:56,890 --> 00:05:00,310 Na siwezi kulinganisha hii kwa kuchukua tu nguzo mbili, moja kutoka Inside Out 88 00:05:00,310 --> 00:05:03,300 na moja kutoka yoyote ya sinema nyingine, na kuona tu 89 00:05:03,300 --> 00:05:06,210 ni wangapi kati yao makala ya mechi. 90 00:05:06,210 --> 00:05:09,660 Hivyo kama mimi kulinganisha Ndani Kati pamoja Marafiki, vizuri, kuna 91 00:05:09,660 --> 00:05:10,910 mambo matatu hapa kwamba mechi hiyo. 92 00:05:10,910 --> 00:05:16,200 Wao ni wote animated, wala wao ni sinema ajabu, na wote wawili 93 00:05:16,200 --> 00:05:18,420 na Mbuga na Req Mbegu. 94 00:05:18,420 --> 00:05:20,420 Ili niweze kuhesabu hadi jinsi mechi nyingi kulikuwa na, 95 00:05:20,420 --> 00:05:22,640 na katika kesi hii kutakuwepo na tatu. 96 00:05:22,640 --> 00:05:26,450 >> Kama mimi kisha kulinganisha Ndani Kati na hebu sema Umri wa Ultron, 97 00:05:26,450 --> 00:05:28,430 Siwezi kuangalia chini orodha na kusema, vizuri, kuna 98 00:05:28,430 --> 00:05:30,140 jambo moja tu kwamba mechi huko. 99 00:05:30,140 --> 00:05:34,560 Wote wawili kupita mtihani Bechtel, hivyo kwamba kinaendelea kuwa alama ya moja. 100 00:05:34,560 --> 00:05:36,770 Na kati ya Ndani Kati na Ant Man, tena mimi 101 00:05:36,770 --> 00:05:41,420 unaweza kulinganisha mstari kwa mstari ni wangapi mambo mechi kati ya watu hao wawili. 102 00:05:41,420 --> 00:05:43,060 Naam, moja ni animated, moja siyo. 103 00:05:43,060 --> 00:05:44,970 Moja ni Marvel movie, moja siyo. 104 00:05:44,970 --> 00:05:47,280 Moja got villain super, wengine hana. 105 00:05:47,280 --> 00:05:49,480 Moja hupita Bechtel mtihani, mtu anashindwa hilo, 106 00:05:49,480 --> 00:05:54,450 lakini wao wote wana Parks na Req Mbegu, hivyo tena, anapata alama ya moja. 107 00:05:54,450 --> 00:05:58,300 >> Hivyo kama mimi walikuwa wanatafuta sinema ambayo yalikuwa sawa na Inside Out, 108 00:05:58,300 --> 00:06:02,170 Mimi naweza kuangalia kwa sinema ambazo zina alama ya juu ndani ya maudhui hii 109 00:06:02,170 --> 00:06:03,952 kuchuja mpango. 110 00:06:03,952 --> 00:06:05,660 Hivyo katika kesi hii, mimi kufikiria Marafiki 111 00:06:05,660 --> 00:06:08,330 kuwa karibu na zaidi uwezekano kuwa kitu 112 00:06:08,330 --> 00:06:13,250 kwamba napenda kutumia fedha kuona kuliko Umri wa Ultron au Ant Man. 113 00:06:13,250 --> 00:06:16,150 >> Yaliyomo hizi msingi mifumo ya kuchuja kutegemea tu 114 00:06:16,150 --> 00:06:18,670 juu ya mali ya sinema, na hivyo mimi 115 00:06:18,670 --> 00:06:21,930 Unaweza kujenga hizi tu kwa kujua kitu kuhusu bidhaa 116 00:06:21,930 --> 00:06:23,500 kwamba nina. 117 00:06:23,500 --> 00:06:26,050 Naweza kutumia aina yoyote ya sifa ya kuwa Ningependa, 118 00:06:26,050 --> 00:06:28,400 na mimi unaweza kujenga zaidi makala tata kwamba 119 00:06:28,400 --> 00:06:33,060 kuhusisha mtihani ngumu zaidi ya ubora kama mimi kwenda pamoja. 120 00:06:33,060 --> 00:06:39,080 Kwa kweli, siwezi hata kuona meza hii si kama moja tuli kitu, 121 00:06:39,080 --> 00:06:43,110 bali kama kuwa vipimo ndani ya kubwa hali nafasi. 122 00:06:43,110 --> 00:06:46,295 Na siwezi kuanza kuzungumza juu ya umbali kati ya sinema mbalimbali. 123 00:06:46,295 --> 00:06:49,300 124 00:06:49,300 --> 00:06:51,050 Haya ni mambo yote kwamba sisi kujua jinsi 125 00:06:51,050 --> 00:06:55,860 kufanya kwa kutumia aina ya miundo data kuwa tumekuwa tayari kuona katika CS50. 126 00:06:55,860 --> 00:06:59,180 Hivyo mimi naweza kufikiria jengo muundo wa data kwa movie. 127 00:06:59,180 --> 00:07:02,390 Kuna struct kwamba nimekuwa yalijengwa aitwaye filamu, 128 00:07:02,390 --> 00:07:04,369 na ina tano viingilio Boolean ndani yake. 129 00:07:04,369 --> 00:07:07,160 Je, ni animated, je, ni Marvel movie, je, ni kuwa villain super, 130 00:07:07,160 --> 00:07:11,047 je, ni kupitisha mtihani Bechdel, na je, kuna Mbuga na Rec Mbegu ndani yake? 131 00:07:11,047 --> 00:07:12,880 Na kila moja ya haya ni muundo wa data kwamba mimi 132 00:07:12,880 --> 00:07:16,330 unaweza kuchukua kwa kuwa filamu fulani. 133 00:07:16,330 --> 00:07:20,090 >> Kisha compute kama mbili sinema ni sawa au la, 134 00:07:20,090 --> 00:07:23,330 nini alama zao ni, mimi naweza kuandika seti ya pseudocode kwamba 135 00:07:23,330 --> 00:07:25,120 inazalisha kazi hiyo. 136 00:07:25,120 --> 00:07:30,100 Hiyo ni, kutokana na baadhi ya movie M1, siwezi kupata movie inayofanana zaidi na hivyo 137 00:07:30,100 --> 00:07:32,430 kwa kufuata pseudocode. 138 00:07:32,430 --> 00:07:37,040 Naona ambayo ni bora akifunga mfumo kuwa Nimepata, 139 00:07:37,040 --> 00:07:39,920 kulinganisha bora kuwa Nimepata. 140 00:07:39,920 --> 00:07:41,890 Kwa kila movie mengine Mimi nina kwenda kupitia, 141 00:07:41,890 --> 00:07:44,920 Mimi itabidi kuweka mechi alama sawa na 0. 142 00:07:44,920 --> 00:07:47,920 Na mimi itabidi kwenda kwa kuwa movie, M1, movie 143 00:07:47,920 --> 00:07:51,500 Mimi kuanza na, mimi itabidi kuangalia kila mmoja na kila kipengele 144 00:07:51,500 --> 00:07:53,650 kwamba wana kuona kama kuna mechi. 145 00:07:53,650 --> 00:07:56,460 Kama kuna mechi, mimi itabidi increment mechi alama. 146 00:07:56,460 --> 00:08:00,480 Na kama katika mwisho wa mechi alama kwamba I have ni bora kuliko ya sasa bora 147 00:08:00,480 --> 00:08:03,310 alama, basi mimi itabidi kumbuka alama kwamba bora, 148 00:08:03,310 --> 00:08:05,820 na hii ni mechi bora kuwa nina. 149 00:08:05,820 --> 00:08:09,450 Mwishoni, chochote movie ni kukaa katika mechi bora, 150 00:08:09,450 --> 00:08:12,580 hiyo ni karibu Nimekuwa na uwezo wa kuja. 151 00:08:12,580 --> 00:08:14,890 Hivyo bidhaa hizi kwa kuzingatia mifumo ya kuchuja, 152 00:08:14,890 --> 00:08:16,900 wote wana muundo huu wa msingi. 153 00:08:16,900 --> 00:08:20,910 Wao wanategemea juu ya bidhaa katika swali na hakuna kitu 154 00:08:20,910 --> 00:08:24,590 kuhusu yoyote ya mapendekezo ya mtumiaji. 155 00:08:24,590 --> 00:08:29,010 >> Utaratibu mwingine kwamba sisi kutumia katika ili kujenga mifumo mapendekezo 156 00:08:29,010 --> 00:08:31,790 inaitwa kuchuja shirikishi. 157 00:08:31,790 --> 00:08:36,520 Kuchuja shirikishi hutegemea juu si sifa ya kitu yenyewe, 158 00:08:36,520 --> 00:08:40,010 lakini jinsi watu, wengine watumiaji, yaani, jinsi wameweza 159 00:08:40,010 --> 00:08:43,370 waliitikia vitu hivi kimoja. 160 00:08:43,370 --> 00:08:48,720 Hivyo kuendelea na movie yangu mfano, Nipate kuupokea kundi la marafiki zangu 161 00:08:48,720 --> 00:08:53,180 na utafiti wao kuhusu kama au si walipenda sinema fulani. 162 00:08:53,180 --> 00:08:56,560 Sasa maeneo mbalimbali kuzalisha taarifa hii kwa njia tofauti. 163 00:08:56,560 --> 00:08:59,630 Unaweza utafiti moja kwa moja yako watumiaji, au unaweza tu 164 00:08:59,630 --> 00:09:03,120 kuona nini wao kuchagua kama uko, kwa mfano Netflix. 165 00:09:03,120 --> 00:09:05,640 Ambayo sinema gani wao kuangalia? 166 00:09:05,640 --> 00:09:08,670 >> Nipate swali baadhi ya yangu marafiki hapa na kujua 167 00:09:08,670 --> 00:09:12,910 kwamba Jason walipenda kila movie aliona, si ajabu huko. 168 00:09:12,910 --> 00:09:15,590 Andy walipenda tu Marafiki na shangazi Man. 169 00:09:15,590 --> 00:09:19,330 Sarah walipenda Ndani Kati na Avengers, kinyume cha Andy. 170 00:09:19,330 --> 00:09:22,200 Na Sam, vizuri, Sam walipenda wote wa sinema superhero, 171 00:09:22,200 --> 00:09:24,960 lakini hakuna hata mmoja sinema animated. 172 00:09:24,960 --> 00:09:30,630 >> Mimi naweza kisha swala kwa baadhi ya mwezi mtu binafsi, baadhi ya mtumiaji mwingine kama mimi 173 00:09:30,630 --> 00:09:34,520 na kuuliza, vizuri, kama mimi walipenda moja ya sinema hizi, 174 00:09:34,520 --> 00:09:38,600 unaweza kufanya utabiri kuhusu ambayo wengine sinema mimi ili kama. 175 00:09:38,600 --> 00:09:41,890 Hiyo ni, kama mimi walipenda Ndani Nje, ambayo sinema nyingine 176 00:09:41,890 --> 00:09:48,460 Mimi ni uwezekano wa pia wanataka kuona kulingana na kile watu sawa alivyofanya? 177 00:09:48,460 --> 00:09:51,640 Yaani, nitakwenda kupitia Mimi itabidi kuchuja kupitia orodha hii 178 00:09:51,640 --> 00:09:54,520 na kupata tu watu ambao pia walipenda 179 00:09:54,520 --> 00:09:57,680 Ndani ya Kati, ambaye kuendana mapendekezo yangu. 180 00:09:57,680 --> 00:10:00,824 Naam, hiyo ina maana kwamba Andy na Sam, wao hawakupenda Ndani Kati, 181 00:10:00,824 --> 00:10:02,240 hivyo mimi si kwenda kufikiria yao. 182 00:10:02,240 --> 00:10:06,130 Mimi nina kwenda kujikwamua yao data kwa kulinganisha hili. 183 00:10:06,130 --> 00:10:09,750 >> Siwezi kisha kuangalia nini Jason na Sarah walidhani na Tally 184 00:10:09,750 --> 00:10:13,780 up yupi kati ya sinema kwamba waliona kwamba mimi si, kama wao walipenda yao 185 00:10:13,780 --> 00:10:15,150 au la. 186 00:10:15,150 --> 00:10:17,820 Mimi nilikuwa tu kuhesabu hadi, hebu sema kura. 187 00:10:17,820 --> 00:10:23,360 Hivyo Marafiki, kwa mfano wanaweza kuwa moja kupiga kura kwa ajili yake, kwa sababu Jason walipenda yake. 188 00:10:23,360 --> 00:10:27,170 Wote Jason na Sarah walipenda Avengers, hivyo ingekuwa kura mbili. 189 00:10:27,170 --> 00:10:30,700 Yasoni tu walipenda Ant Man, hivyo itakuwa kupata kura moja. 190 00:10:30,700 --> 00:10:34,870 Hivyo kama mimi alikuwa na kisha kupendekeza kwa mwenyewe yupi kati ya hawa sinema 191 00:10:34,870 --> 00:10:41,470 Nipate kuwa na uwezo mkubwa kwa kuangalia, napenda Una kuchagua Umri wa Ultron: Avengers. 192 00:10:41,470 --> 00:10:44,490 >> Hivyo kwa yoyote ya hizi mifumo, sasa mimi nina kutumia 193 00:10:44,490 --> 00:10:49,260 takwimu ambazo ilitokana si kuhusu movie yenyewe, lakini kuhusu mapendekezo 194 00:10:49,260 --> 00:10:51,960 kutoka kwa watumiaji wengine. 195 00:10:51,960 --> 00:10:54,150 Hii ina baadhi ya matatizo bila shaka. 196 00:10:54,150 --> 00:10:55,920 Nini kama huna watumiaji wengine yoyote? 197 00:10:55,920 --> 00:10:58,770 Naam, kwamba wito tatizo startup. 198 00:10:58,770 --> 00:11:03,760 Unaweza kuwa na baadhi wingi wa data kabla uko 199 00:11:03,760 --> 00:11:07,560 uwezo wa kuanza kufanya mapendekezo haya. 200 00:11:07,560 --> 00:11:10,940 Hasara ya ni mara moja kuanza kukusanya takwimu, 201 00:11:10,940 --> 00:11:13,870 kama unaweza kukusanya zaidi na data zaidi na zaidi, 202 00:11:13,870 --> 00:11:17,850 utasikia kupata bora na bora na mapendekezo bora. 203 00:11:17,850 --> 00:11:21,650 >> Sasa tunaweza kutafsiri hii katika kanuni vilevile. 204 00:11:21,650 --> 00:11:23,860 Tunaweza kufafanua tofauti aina ya muundo, 205 00:11:23,860 --> 00:11:25,720 katika kesi hii tutaweza simu yake mtumiaji. 206 00:11:25,720 --> 00:11:30,970 Na ni got sifa kuhusu ambayo sinema mtumiaji huyu walipenda. 207 00:11:30,970 --> 00:11:34,560 Je, wao kama Inside Out, Marafiki, Avengers, na Ant Man. 208 00:11:34,560 --> 00:11:36,660 Tunaweza kisha kuzalisha baadhi pseudocode kufuata 209 00:11:36,660 --> 00:11:39,460 utaratibu huo kwamba mimi kutumika kabla. 210 00:11:39,460 --> 00:11:43,460 Hiyo ni, kutokana na hasa user x, hebu kupendekeza movie 211 00:11:43,460 --> 00:11:46,107 kwamba x ili kama. 212 00:11:46,107 --> 00:11:47,940 Tunaweza kwenda kwa njia na kwa wote wa sinema, 213 00:11:47,940 --> 00:11:51,410 tunaweza initialize alama kwa kuwa filamu kuwa 0. 214 00:11:51,410 --> 00:11:54,080 Na kisha tunaweza kupata yote ya watumiaji wengine ambao 215 00:11:54,080 --> 00:11:57,630 wana matakwa sawa na x. 216 00:11:57,630 --> 00:11:59,990 Na kisha kwa kila movie kwamba walipenda, 217 00:11:59,990 --> 00:12:02,340 tutaweza increment alama ya kwamba movie. 218 00:12:02,340 --> 00:12:05,010 Kwa namna yoyote movie katika mwisho ina alama ya juu, 219 00:12:05,010 --> 00:12:07,600 hiyo ni moja mimi lazima kupendekeza. 220 00:12:07,600 --> 00:12:09,890 >> Hakuna hata hii ni kweli wasio na cheo. 221 00:12:09,890 --> 00:12:11,600 Hakuna wa hii ni changamoto. 222 00:12:11,600 --> 00:12:15,810 Hizi ni algorithms yote ya msingi kwamba unaweza kutekeleza leo. 223 00:12:15,810 --> 00:12:20,050 >> Sasa na mifumo MKUBALIAJI halisi, wewe kukimbia katika baadhi ya matatizo. 224 00:12:20,050 --> 00:12:23,300 Nini kama kuna hakuna mtu ambaye mechi hasa mapendekezo yako? 225 00:12:23,300 --> 00:12:27,170 Nini kama kuna watumiaji ambao ni hasa mapendekezo yako, 226 00:12:27,170 --> 00:12:30,480 lakini kisha uadilifu kwa kiasi kikubwa kutokana na kile wewe kama? 227 00:12:30,480 --> 00:12:36,210 Mimi kama classic Godzilla sinema, lakini mke wangu hana. 228 00:12:36,210 --> 00:12:39,430 Mimi kama kuangalia yao, yangu Netflix akaunti ina kwao. 229 00:12:39,430 --> 00:12:41,800 Yake haina. 230 00:12:41,800 --> 00:12:45,230 Kile kinachotokea wakati sisi kuanza kuchanganya data kama hii? 231 00:12:45,230 --> 00:12:47,690 Hizi ni changamoto zote uweze kushinda, 232 00:12:47,690 --> 00:12:51,900 wao tu kuchukua kidogo ngumu zaidi algorithms. 233 00:12:51,900 --> 00:12:56,420 >> Sasa katika ulimwengu wa kweli, ambayo ni kweli kazi, 234 00:12:56,420 --> 00:12:59,980 tunatumia maudhui ya msingi ya kuchuja au tunatumia kuchuja shirikishi? 235 00:12:59,980 --> 00:13:01,910 Na jibu ni sisi kutumia wote wawili. 236 00:13:01,910 --> 00:13:06,350 Karibu wote wa watumiaji kubwa katika kesi hiyo, Amazon, Facebook, Netflix, 237 00:13:06,350 --> 00:13:11,200 Pandora, wote kutumia mchanganyiko wa hizi mbalimbali mifumo mapendekezo. 238 00:13:11,200 --> 00:13:16,520 Na wakati sisi kuchanganya uchaguzi kutoka kila mmoja, sisi kuwaita mifumo ya mseto. 239 00:13:16,520 --> 00:13:20,750 Wao kwa namna fulani hutegemea juu ya sifa za kitu yenyewe, 240 00:13:20,750 --> 00:13:24,710 na katika baadhi ya njia wao hutegemea juu ya matakwa ya watumiaji wengine. 241 00:13:24,710 --> 00:13:28,120 Mifumo hii ya mseto, wao uko biashara kubwa, 242 00:13:28,120 --> 00:13:30,830 na wao uko nini sasa leo. 243 00:13:30,830 --> 00:13:32,839 >> Hivyo shukrani sana kwa kujiunga na mimi. 244 00:13:32,839 --> 00:13:35,380 Natumaini wameweza kujipatia kidogo kidogo ya uelewa wa nini 245 00:13:35,380 --> 00:13:37,430 hufanya mifumo hii kazi. 246 00:13:37,430 --> 00:13:41,980 Next wakati wewe ni online, kukumbuka kwamba si tu wewe kushawishi uchaguzi wako, 247 00:13:41,980 --> 00:13:44,680 lakini uwezekano wa mtu mwingine ni kama vile. 248 00:13:44,680 --> 00:13:46,480 Asante tena. 249 00:13:46,480 --> 00:13:47,186