検索キーワードをPythonのScikit learnでクラスタリングした話 〜機械学習を使って自然検索に強いサイトを作る〜
by
soheiyagi
Link
Embed
Share
Beginning
This slide
Copy link URL
Copy link URL
Copy iframe embed code
Copy iframe embed code
Copy javascript embed code
Copy javascript embed code
Share
Tweet
Share
Tweet
Slide 1
Slide 1 text
ػցֶशΛͬͯࣗવݕࡧʹڧ͍αΠτΛ࡞Δ ݕࡧΩʔϫʔυΛTDJLJUMFBSOͰ ΫϥελϦϯάͨ͠ 406 ˏTPIFJZBHJ TPIFJZBHJ
Slide 2
Slide 2 text
ࣗݾհ ͜Μͳਓʹฉ͍ͯཉ͍͠ Ͳ͜ͰΫϥελϦϯάΛ͔ͬͨ ίʔυͲΜͳײ͡ ·ͱΊ ࣍
Slide 3
Slide 3 text
ࣗݾհ גࣜձࣾιදऔక TPVDPKQ ɾίϯςϯπϚʔέςΟϯά ɾ8FCࠂӡ༻ ɾϚʔέςΟϯάπʔϧ࡞ ຊொΦʔϓϯιʔεϥϘӡӦ ɾΤϯδχΞͷίϛϡχςΟεϖʔε େࡕ1ZUIPOͷձɹΦʔΨφΠβʔ தখاۀிϛϥαϙઐՈݣɹొઐՈ 8FCࠂਓࡐϏδωεͷӦۀ͔ΒࣄΛ͡Ίɺ 8FCӡ༻ࠂӡ༻ΛΓͭͭɺϓϩάϥϜΛॻ͘ਓ ίʔυ͕ॻ͚ΔϚʔέολʔ !TPIFJZBHJ TPIFJZBHJ 4PIFJ:BHJʗീฏ
Slide 4
Slide 4 text
ࣗݾհ גࣜձࣾιදऔక TPVDPKQ ɾίϯςϯπϚʔέςΟϯά ɾ8FCࠂӡ༻ ɾϚʔέςΟϯάπʔϧ࡞ ຊொΦʔϓϯιʔεϥϘӡӦ ɾΤϯδχΞͷίϛϡχςΟεϖʔε େࡕ1ZUIPOͷձɹΦʔΨφΠβʔ தখاۀிϛϥαϙઐՈݣɹొઐՈ 8FCࠂਓࡐϏδωεͷӦۀ͔ΒࣄΛ͡Ίɺ 8FCӡ༻ࠂӡ༻ΛΓͭͭɺϓϩάϥϜΛॻ͘ਓ ίʔυ͕ॻ͚ΔϚʔέολʔ !TPIFJZBHJ TPIFJZBHJ 4PIFJ:BHJʗീฏ ࠷ۙͷझຯ ےτϨͰ͢ɻ
Slide 5
Slide 5 text
ࣗݾհ ͜Μͳਓʹฉ͍ͯཉ͍͠ Ͳ͜ͰΫϥελϦϯάΛ͔ͬͨ ίʔυͲΜͳײ͡ ·ͱΊ ࣍
Slide 6
Slide 6 text
͜Μͳਓʹฉ͍ͯཉ͍͠ 8FCαΠτΛӡӦ͍ͯ͠Δਓ ϒϩάΛӡӦ͍ͯ͠Δਓ &$αΠτΛӡӦ͍ͯ͠Δਓ ɾɾɾ 8FCαΠτ͔Β Կ͔͠ΒͷՌ͕ཉ͍͠ਓ
Slide 7
Slide 7 text
ࣗݾհ ͜Μͳਓʹฉ͍ͯཉ͍͠ Ͳ͜ͰΫϥελϦϯάΛ͔ͬͨ ίʔυͲΜͳײ͡ ·ͱΊ ࣍
Slide 8
Slide 8 text
Ͳ͜ͰΫϥελϦϯάΛ͔ͬͨ 8FCαΠτઃܭͷྲྀΕ αδΣετϫʔυͷऔಘ ʢϢʔβʔχʔζ͕͋ΔΩʔϫʔυ܈ʣ ҙຯͷ͋Δմʹάϧʔϐϯά άϧʔϐϯάΛݩʹαΠτઃܭ
Slide 9
Slide 9 text
Ͳ͜ͰΫϥελϦϯάΛ͔ͬͨ 8FCαΠτઃܭͷྲྀΕ αδΣετϫʔυͷऔಘ ʢϢʔβʔχʔζ͕͋ΔΩʔϫʔυ܈ʣ ҙຯͷ͋Δմʹάϧʔϐϯά άϧʔϐϯάΛݩʹαΠτઃܭ
Slide 10
Slide 10 text
αδΣετϫʔυͷऔಘ Ϣʔβʔχʔζ͕͋ΔΩʔϫʔυ܈ΛूΊΔ
Slide 11
Slide 11 text
αδΣετϫʔυͷऔಘ Πϊϕʔγϣϯࣄྫ Πϊϕʔγϣϯࣄྫຊ Πϊϕʔγϣϯࣄྫ࠷ۙ Πϊϕʔγϣϯࣄྫ Πϊϕʔγϣϯࣄྫւ֎ ΠϊϕʔγϣϯࣄྫJQIPOF Πϊϕʔγϣϯࣄྫ࠷৽ Πϊϕʔγϣϯࣄྫۙ ΠϊϕʔγϣϯࣄྫΞϝϦΧ ΠϊϕʔγϣϯࣄྫΥʔΫϚϯ ΤίγεςϜΠϊϕʔγϣϯࣄྫ ӦۀΠϊϕʔγϣϯࣄྫ ΠϊϕʔγϣϯΦϑΟεࣄྫ ΦʔϓϯΠϊϕʔγϣϯࣄྫ ΦʔϓϯΠϊϕʔγϣϯࣄྫຊ ΦʔϓϯσʔλΠϊϕʔγϣϯࣄྫ େࡕΨεΠϊϕʔγϣϯࣄྫ Πϊϕʔγϣϯձࣾࣄྫ Πϊϕʔγϣϯڥࣄྫ ՁΠϊϕʔγϣϯࣄྫ ֶੜΠϊϕʔγϣϯࣄྫ Πϊϕʔγϣϯࣄྫاۀ Πϊϕʔγϣϯۚ༥ࣄྫ େاۀΠϊϕʔγϣϯࣄྫ ٕज़Πϊϕʔγϣϯࣄྫ ۀքΠϊϕʔγϣϯࣄྫ ΠϊϕʔγϣϯࣄྫΈ߹Θͤ ݚڀ։ൃΠϊϕʔγϣϯࣄྫ Πϊϕʔγϣϯܦࡁࣄྫ ΠϊϕʔγϣϯܦӦࣄྫ খചۀΠϊϕʔγϣϯࣄྫ খചΠϊϕʔγϣϯࣄྫ ࢠڙΠϊϕʔγϣϯࣄྫ খചۀΠϊϕʔγϣϯࣄྫ ΠϊϕʔγϣϯࣄྫαʔϏε ࢈ֶ࿈ܞΠϊϕʔγϣϯࣄྫ αʔϏεۀΠϊϕʔγϣϯࣄྫ Πϊϕʔγϣϯࣄྫू Πϊϕʔγϣϯࣄྫࣦഊ Πϊϕʔγϣϯࣄྫ Πϊϕʔγϣϯ৽݁߹ࣄྫ ΤίγεςϜΠϊϕʔγϣϯࣄྫ ࣾձ՝Πϊϕʔγϣϯࣄྫ ࣾΠϊϕʔγϣϯࣄྫ ΠϊϕʔγϣϯδϨϯϚࣄྫ Πϊϕʔγϣϯ࣋ଓతࣄྫ Πϊϕʔγϣϯࣄྫੈք Πϊϕʔγϣϯࣄྫ ۀΠϊϕʔγϣϯࣄྫ ଟ༷ੑΠϊϕʔγϣϯࣄྫ Πϊϕʔγϣϯ৫ࣄྫ Πϊϕʔγϣϯग़ࣄྫ ιʔγϟϧΠϊϕʔγϣϯࣄྫຊ ൃΠϊϕʔγϣϯࣄྫ ଟ༷ੑΠϊϕʔγϣϯࣄྫ େاۀΠϊϕʔγϣϯࣄྫ μΠόʔγςΟΠϊϕʔγϣϯࣄྫ தখاۀΠϊϕʔγϣϯࣄྫ ΠϊϕʔγϣϯνʔϜࣄྫ Πϊϕʔγϣϯࣝࣄྫ ҬΠϊϕʔγϣϯࣄྫ ΠϊϕʔγϣϯࣄྫσβΠϯࢥߟ σβΠϯΠϊϕʔγϣϯࣄྫ σδλϧΠϊϕʔγϣϯࣄྫ ഁյతΠϊϕʔγϣϯࣄྫ ඇ࿈ଓΠϊϕʔγϣϯࣄྫ ཱΠϊϕʔγϣϯࣄྫ ϏδωεϞσϧΠϊϕʔγϣϯࣄྫ ϏδωεΠϊϕʔγϣϯࣄྫ ϏοάσʔλΠϊϕʔγϣϯࣄྫ ࢜ϑΠϧϜΠϊϕʔγϣϯࣄྫ ࢜௨ϑΟʔϧυΠϊϕʔγϣϯࣄྫ ࢜௨Πϊϕʔγϣϯࣄྫ ϓϩηεΠϊϕʔγϣϯࣄྫ ϓϩμΫτΠϊϕʔγϣϯࣄྫ ϔϧεέΞΠϊϕʔγϣϯࣄྫ Πϊϕʔγϣϯࣄྫຊ ΠϊϕʔγϣϯࣄྫϚΠΫϩιϑτ ϚʔέςΟϯάΠϊϕʔγϣϯࣄྫ ϢʔβʔΠϊϕʔγϣϯࣄྫ ϦόʔεΠϊϕʔγϣϯࣄྫ ΞʔΩςΫνϟϧΠϊϕʔγϣϯࣄྫ ࢈ֶ࿈ܞΠϊϕʔγϣϯࣄྫ QHΠϊϕʔγϣϯࣄྫ Πϊϕʔγϣϯࣄྫ NΠϊϕʔγϣϯࣄྫ Πϊϕʔγϣϯͭͷػձࣄྫ λʔήοτΩʔϫʔυʮΠϊϕʔγϣϯࣄྫʯ݅
Slide 12
Slide 12 text
ҙຯͷ͋Δմʹάϧʔϐϯά ɾཱΠϊϕʔγϣϯࣄྫ ɾ࢜ϑΠϧϜΠϊϕʔγϣϯࣄྫ ɾେࡕΨεΠϊϕʔγϣϯࣄྫ ɾμΠόʔγςΟΠϊϕʔγϣϯࣄྫ ɾଟ༷ੑΠϊϕʔγϣϯࣄྫ ɾଟ༷ੑΠϊϕʔγϣϯࣄྫ ɾΠϊϕʔγϣϯδϨϯϚࣄྫ ɾΠϊϕʔγϣϯ࣋ଓతࣄྫ ɾഁյతΠϊϕʔγϣϯࣄྫ اۀࣄྫ ଟ༷ੑ δϨϯϚ ࢹͰ͍ͬͯͨͷΛɺ ػցֶशͰάϧʔϓ͚Λߦͬͨ
Slide 13
Slide 13 text
࣮ࡍͷਫ਼ ΠϊϕʔγϣϯܦӦࣄྫ Πϊϕʔγϣϯܦࡁࣄྫ Πϊϕʔγϣϯࣄྫ ΠϊϕʔγϣϯࣄྫΥʔΫϚϯ ϓϩηεΠϊϕʔγϣϯࣄྫ ඇ࿈ଓΠϊϕʔγϣϯࣄྫ Πϊϕʔγϣϯձࣾࣄྫ ΠϊϕʔγϣϯࣄྫΞϝϦΧ ΠϊϕʔγϣϯࣄྫαʔϏε Πϊϕʔγϣϯࣄྫاۀ Πϊϕʔγϣϯࣄྫू Πϊϕʔγϣϯࣄྫۙ Πϊϕʔγϣϯࣄྫੈք Πϊϕʔγϣϯࣄྫ ϓϩμΫτΠϊϕʔγϣϯࣄྫ ٕज़Πϊϕʔγϣϯࣄྫ ࢠڙΠϊϕʔγϣϯࣄྫ μΠόʔγςΟΠϊϕʔγϣϯࣄྫ ଟ༷ੑΠϊϕʔγϣϯࣄྫ ଟ༷ੑΠϊϕʔγϣϯࣄྫ Πϊϕʔγϣϯࣄྫࣦഊ Πϊϕʔγϣϯ৫ࣄྫ ֶੜΠϊϕʔγϣϯࣄྫ Πϊϕʔγϣϯࣄྫւ֎ ΦʔϓϯΠϊϕʔγϣϯࣄྫ ΦʔϓϯΠϊϕʔγϣϯࣄྫຊ ΠϊϕʔγϣϯࣄྫσβΠϯࢥߟ σβΠϯΠϊϕʔγϣϯࣄྫ Πϊϕʔγϣϯࣄྫ Πϊϕʔγϣϯࣄྫ࠷ۙ Πϊϕʔγϣϯࣄྫ࠷৽ Πϊϕʔγϣϯࣄྫຊ Πϊϕʔγϣϯࣄྫ αʔϏεۀΠϊϕʔγϣϯࣄྫ ϏδωεΠϊϕʔγϣϯࣄྫ ۀքΠϊϕʔγϣϯࣄྫ ۀΠϊϕʔγϣϯࣄྫ Πϊϕʔγϣϯڥࣄྫ Πϊϕʔγϣϯۚ༥ࣄྫ ιʔγϟϧΠϊϕʔγϣϯࣄྫຊ ϏδωεϞσϧΠϊϕʔγϣϯࣄྫ ΠϊϕʔγϣϯδϨϯϚࣄྫ Πϊϕʔγϣϯ࣋ଓతࣄྫ ഁյతΠϊϕʔγϣϯࣄྫ খചΠϊϕʔγϣϯࣄྫ খചۀΠϊϕʔγϣϯࣄྫ খചۀΠϊϕʔγϣϯࣄྫ ࢜௨Πϊϕʔγϣϯࣄྫ ࢜௨ϑΟʔϧυΠϊϕʔγϣϯࣄྫ NΠϊϕʔγϣϯࣄྫ QHΠϊϕʔγϣϯࣄྫ ΞʔΩςΫνϟϧΠϊϕʔγϣϯࣄྫ ΠϊϕʔγϣϯΦϑΟεࣄྫ ΠϊϕʔγϣϯνʔϜࣄྫ ΠϊϕʔγϣϯࣄྫϚΠΫϩιϑτ Πϊϕʔγϣϯࣄྫຊ Πϊϕʔγϣϯग़ࣄྫ Πϊϕʔγϣϯࣝࣄྫ ΦʔϓϯσʔλΠϊϕʔγϣϯࣄྫ σδλϧΠϊϕʔγϣϯࣄྫ ϏοάσʔλΠϊϕʔγϣϯࣄྫ ϔϧεέΞΠϊϕʔγϣϯࣄྫ ϢʔβʔΠϊϕʔγϣϯࣄྫ ϦόʔεΠϊϕʔγϣϯࣄྫ ӦۀΠϊϕʔγϣϯࣄྫ ݚڀ։ൃΠϊϕʔγϣϯࣄྫ ࢈ֶ࿈ܞΠϊϕʔγϣϯࣄྫ ࢈ֶ࿈ܞΠϊϕʔγϣϯࣄྫ ࣾձ՝Πϊϕʔγϣϯࣄྫ ࣾΠϊϕʔγϣϯࣄྫ ൃΠϊϕʔγϣϯࣄྫ େࡕΨεΠϊϕʔγϣϯࣄྫ ҬΠϊϕʔγϣϯࣄྫ ཱΠϊϕʔγϣϯࣄྫ ࢜ϑΠϧϜΠϊϕʔγϣϯࣄྫ
Slide 14
Slide 14 text
࣮ࡍͷਫ਼ ΤίγεςϜΠϊϕʔγϣϯࣄྫ ΤίγεςϜΠϊϕʔγϣϯࣄྫ େاۀΠϊϕʔγϣϯࣄྫ େاۀΠϊϕʔγϣϯࣄྫ தখاۀΠϊϕʔγϣϯࣄྫ ΠϊϕʔγϣϯࣄྫJQIPOF ΠϊϕʔγϣϯࣄྫΈ߹Θͤ Πϊϕʔγϣϯ৽݁߹ࣄྫ ϚʔέςΟϯάΠϊϕʔγϣϯࣄྫ ՁΠϊϕʔγϣϯࣄྫ Πϊϕʔγϣϯͭͷػձࣄྫ ·ͩ·ͩਫ਼্͛Δඞཁ͋Δ͕ɺ ࢹͰҰ͔ΒΔΑΓஅવ࣌ؒॖ͕࣮ݱ
Slide 15
Slide 15 text
άϧʔϐϯάΛݩʹαΠτઃܭ اۀࣄྫ ଟ༷ੑ δϨϯϚ Πϊϕʔγϣϯࣄྫͷϖʔδ Ϣʔβʔχʔζͷ͋ΔΩʔϫʔυΛͬͯ αΠτઃܭ͍ͯ͘͠ͷͰݕࡧʹڧ͍αΠτઃܭʹͳΔ ɾɾɾ
Slide 16
Slide 16 text
ຊொΦʔϓϯιʔεϥϘ8FCαΠτͰެ։࣮ݧ తɺ*5ษڧձʹڵຯΛ࣋ͬͯΒ͍ΞΫγϣϯΛىͯ͜͠Β͏ࣄ IUUQTIPNNBDIJPQFOTPVSDFMBCHJUIVCJPɹ
Slide 17
Slide 17 text
݄͔Β αΠτઃܭͨ͠هࣄΛՃ ຊொΦʔϓϯιʔεϥϘ8FCαΠτͰެ։࣮ݧ
Slide 18
Slide 18 text
ຊொΦʔϓϯιʔεϥϘ8FCαΠτͰެ։࣮ݧ Ωʔϫʔυ ॱҐ ݕࡧϘϦϡʔϜ ϓϩάϥϛϯάษڧձॳ৺ऀ ϓϩάϥϛϯάษڧձॳ৺ऀ JUษڧձॳ৺ऀ JUॳ৺ऀηϛφʔ VEFNZQZUIPO͓͢͢Ί ίϯύεษڧձ JUษڧձ JUษڧձ DPNQBTTษڧձ ษڧձαΠτ ΤϯδχΞษڧձ ษڧձJU EPUTษڧձ ษڧձɹJU JUษڧձΧϨϯμʔ ΤϯδχΞษڧձ JUษڧॳ৺ऀ ໊ݹQZUIPO QZUIPOॳ৺ऀษڧձ Ωʔϫʔυ ॱҐ ݕࡧϘϦϡʔϜ JUษڧձେࡕ େࡕJUษڧձ ϓϩάϥϛϯάษڧձ େࡕJUษڧձ JUษڧձେࡕ QZUIPOษڧձ౦ژ େࡕษڧձJU ԬJUษڧձ ϓϩάϥϛϯάษڧձ SVCZؔ େࡕJUษڧձ ໊ݹษڧձ QZUIPOηϛφʔ QZUIPOVEFNZ VEFNZQZUIPO JUษڧ ૂͬͨΩʔϫʔυͰͷ্Ґදࣔͱ$7Λ֫ಘ ˞݄࣌ɹ"ISFGTௐ
Slide 19
Slide 19 text
˞݄࣌ɹ"ISFGTௐ ຊொΦʔϓϯιʔεϥϘ8FCαΠτͰެ։࣮ݧ Ωʔϫʔυ ॱҐ ݕࡧϘϦϡʔϜ ϓϩάϥϛϯάษڧձॳ৺ऀ ϓϩάϥϛϯάษڧձॳ৺ऀ JUษڧձॳ৺ऀ JUॳ৺ऀηϛφʔ VEFNZQZUIPO͓͢͢Ί ίϯύεษڧձ JUษڧձ JUษڧձ DPNQBTTษڧձ ษڧձαΠτ ΤϯδχΞษڧձ ษڧձJU EPUTษڧձ ษڧձɹJU JUษڧձΧϨϯμʔ ΤϯδχΞษڧձ JUษڧॳ৺ऀ ໊ݹQZUIPO QZUIPOॳ৺ऀษڧձ Ωʔϫʔυ ॱҐ ݕࡧϘϦϡʔϜ JUษڧձେࡕ େࡕJUษڧձ ϓϩάϥϛϯάษڧձ େࡕJUษڧձ JUษڧձେࡕ QZUIPOษڧձ౦ژ େࡕษڧձJU ԬJUษڧձ ϓϩάϥϛϯάษڧձ SVCZؔ େࡕJUษڧձ ໊ݹษڧձ QZUIPOηϛφʔ QZUIPOVEFNZ VEFNZQZUIPO JUษڧ ࠓճͷςʔϚʹؔͳ͍Ͱ͕͢ɺ ݕࡧΩʔϫʔυͷϘϦϡʔϜͱ $73ͷ૬ؔແ͍ͨΊɺ ࣮ࡍʹ$7ʹ͍ۙΩʔϫʔυͰ ্ҐදࣔͰ͖Δ͜ͱ͕ॏཁ ˞$73ɹίϯόʔδϣϯʢʣ ˞$7ɹίϯόʔδϣϯ
Slide 20
Slide 20 text
ࣗݾհ ͜Μͳਓʹฉ͍ͯཉ͍͠ Ͳ͜ͰΫϥελϦϯάΛ͔ͬͨ ίʔυͲΜͳײ͡ ·ͱΊ ࣍
Slide 21
Slide 21 text
ίʔυͲΜͳײ͡ # -*- coding: utf-8 -*- import pandas as pd from sklearn.cluster import KMeans ίʔυൈਮ pred = KMeans( n_clusters=int(query_len/5), init='k-means++' ).fit_predict(cust_array)
Slide 22
Slide 22 text
# -*- coding: utf-8 -*- import pandas as pd from sklearn.cluster import KMeans ίʔυൈਮ pred = KMeans( n_clusters=int(query_len/5), init='k-means++' ).fit_predict(cust_array) ,.FBOTΛར༻ ίʔυͲΜͳײ͡
Slide 23
Slide 23 text
LNFBOT๏ͷΠϝʔδ ΫϥελϦϯάͷख๏ͷͭͰΑ͘ΘΕ͍ͯΔ
Slide 24
Slide 24 text
Ϋϥελͷத৺ΛϥϯμϜʹܾΊΔ ݸΫϥελ LNFBOT๏ͷΠϝʔδ
Slide 25
Slide 25 text
֤σʔλΛۙ͘ͷΫϥελத৺ʹूΊΔ LNFBOT๏ͷΠϝʔδ
Slide 26
Slide 26 text
֤σʔλͷฏۉͰ৽͍͠த৺Λઃఆ͢Δ LNFBOT๏ͷΠϝʔδ
Slide 27
Slide 27 text
֤σʔλΛۙ͘ͷΫϥελத৺ʹूΊΔ LNFBOT๏ͷΠϝʔδ
Slide 28
Slide 28 text
֤σʔλͷฏۉͰ৽͍͠த৺Λઃఆ͢Δ LNFBOT๏ͷΠϝʔδ
Slide 29
Slide 29 text
֤σʔλΛۙ͘ͷΫϥελத৺ʹूΊΔ LNFBOT๏ͷΠϝʔδ
Slide 30
Slide 30 text
த৺͕ಈ͔ͳ͘ͳΔ·Ͱ܁Γฦ͢ LNFBOT๏ͷΠϝʔδ
Slide 31
Slide 31 text
# -*- coding: utf-8 -*- import pandas as pd from sklearn.cluster import KMeans ίʔυൈਮ pred = KMeans( n_clusters=int(query_len/5), init='k-means++' ).fit_predict(cust_array) Ϋϥελ ίʔυͲΜͳײ͡
Slide 32
Slide 32 text
# -*- coding: utf-8 -*- import pandas as pd from sklearn.cluster import KMeans ίʔυൈਮ pred = KMeans( n_clusters=int(query_len/5), init='k-means++' ).fit_predict(cust_array) αδΣετΩʔϫʔυΛ ͰׂͬͨࣈΛઃఆ ˞ҙͷࣈ ίʔυͲΜͳײ͡
Slide 33
Slide 33 text
# -*- coding: utf-8 -*- import pandas as pd from sklearn.cluster import KMeans ίʔυൈਮ pred = KMeans( n_clusters=int(query_len/5), init='k-means++' ).fit_predict(cust_array) ॳظԽͷઃఆ ίʔυͲΜͳײ͡
Slide 34
Slide 34 text
# -*- coding: utf-8 -*- import pandas as pd from sklearn.cluster import KMeans ίʔυൈਮ pred = KMeans( n_clusters=int(query_len/5), init='k-means++' ).fit_predict(cust_array) ֤σʔλʹର͢Δ Ϋϥελ൪߸Λฦ͢ ίʔυͲΜͳײ͡
Slide 35
Slide 35 text
# -*- coding: utf-8 -*- import pandas as pd from sklearn.cluster import KMeans ίʔυൈਮ pred = KMeans( n_clusters=int(query_len/5), init='k-means++' ).fit_predict(cust_array) <<> <> <> <>> ֤ΩʔϫʔυΛϕΫτϧԽͯ͠pU@QSFEJDUʹ͢ Ωʔϫʔυ< > Ωʔϫʔυ< > ίʔυͲΜͳײ͡
Slide 36
Slide 36 text
# -*- coding: utf-8 -*- import pandas as pd from sklearn.cluster import KMeans ίʔυൈਮ pred = KMeans( n_clusters=int(query_len/5), init='k-means++' ).fit_predict(cust_array) < > ֤ΩʔϫʔυͷΫϥελ൪߸͕ฦͬͯ͘Δ Ωʔϫʔυɿ Ωʔϫʔυɿ̍̒ ίʔυͲΜͳײ͡
Slide 37
Slide 37 text
ࣗݾհ ͜Μͳਓʹฉ͍ͯཉ͍͠ Ͳ͜ͰΫϥελϦϯάΛ͔ͬͨ ίʔυͲΜͳײ͡ ·ͱΊ ࣍
Slide 38
Slide 38 text
·ͱΊ ɾαΠτઃܭͷαδΣετΩʔϫʔυͷάϧʔϐϯάͷ࡞ ۀ͕ѹతʹָʹͳͬͨ ɾαδΣετΩʔϫʔυ͕ଟ͘ͳΔͱɺLNFBOTͷܭ ࢉ͕࣌ؒരൃ͢Δ ɾڭࢣແ͠ͷػցֶशͷͨΊɺ݁ՌࢹͰଥ͔Ͳ͏ ͔ͷஅ͕ඞཁ ɾࣗવݴޠॲཧΛҰॹʹษڧ͍ͨ͠ਓ͍Ε͔͚ͯ͘ ͍ͩ͞ ɾࣗવݕࡧʹڧ͍αΠτΛ࡞Γ͍ͨਓ͔͚ͯͩ͘͞ ͍