åå¿è åãã®ããŒã¿ãµã€ãšã³ã¹
1. ã»ã³ãã¡ã³ãåæïŒããã¹ãã«ããæ°ååæïŒ
ãœãŒã¹ã³ãŒãã䜿çšããŠããŒã¿ãµã€ãšã³ã¹ãããžã§ã¯ãã®å®å
šãªå®è£
ã衚瀺ããŸã-
ææ åæã¯ãè¯å®çãŸãã¯åŠå®çãªææ ãæèŠãèå¥ããããã®åèªã®åæã§ãã ããã¯åé¡ã®äžçš®ã§ãã¯ã©ã¹ã¯äºå€ (ããžãã£ããšãã¬ãã£ã) ãŸãã¯è€æ° (幞ããæããæ²ãã¿ãæå°æªãªã©) ã«ãªããŸãã ãã®ããŒã¿ ãµã€ãšã³ã¹ ãããžã§ã¯ãã R ã§å®è£ ãããjaneaustenRãããã±ãŒãžã®ããŒã¿ã»ããã䜿çšããŸãã AFINNãbingãloughran ãªã©ã®æ±çšèŸæžã䜿çšããå éšçµåãå®è¡ããæåŸã«çµæã衚瀺ããã¯ãŒã ã¯ã©ãŠããäœæããŸãã
èšèªïŒ R
ããŒã¿ã»ãã/ããã±ãŒãž: ãžã§ãŒã³ãªãŒã¹ãã³
ãã®èšäºã¯ EDISON Software ã®ãµããŒããåããŠç¿»èš³ãããŸããããã«ããã©ã³ãã¹ãã¢åãã®ä»®æ³è©Šç宀ãäœæããŸã ãšãã¹ããœãããŠã§ã¢ .
2. ãã§ã€ã¯ãã¥ãŒã¹ã®æ€åº
åå¿è
åãã®ããŒã¿ ãµã€ãšã³ã¹ ãããžã§ã¯ãã«åãçµãã§ãã¹ãã«ã次ã®ã¬ãã«ã«åŒãäžããŸããã-
ãã§ã€ã¯ãã¥ãŒã¹ãšã¯ãæ¿æ²»çç®æšãéæããããã«ãœãŒã·ã£ã«ã¡ãã£ã¢ããã®ä»ã®ãªã³ã©ã€ã³ã¡ãã£ã¢ãéããŠæ¡æ£ããã誀ã£ãæ
å ±ã§ãã ãã®ããŒã¿ ãµã€ãšã³ã¹ ãããžã§ã¯ãã®ã¢ã€ãã¢ã§ã¯ãPython ã䜿çšããŠããã¥ãŒã¹ãæ¬ç©ãåœç©ããæ£ç¢ºã«å€æã§ããã¢ãã«ãæ§ç¯ããŸãã TfidfVectorizer ãäœæããPassiveAggressiveClassifier ã䜿çšããŠãã¥ãŒã¹ããæ¬ç©ããšãåœç©ãã«åé¡ããŸãã 7796Ã4 圢ç¶ããŒã¿ã»ããã䜿çšãããã¹ãŠã Jupyter Lab ã§è¡ããŸãã
èšèªïŒ Python
ããŒã¿ã»ãã/ããã±ãŒãž: ãã¥ãŒã¹.csv
3. ããŒãã³ãœã³ç ã®æ€åº
ããŒã¿ãµã€ãšã³ã¹ãããžã§ã¯ãã®ã¢ã€ãã¢ã«åãçµãã§åé²ããŸã-
ç§ãã¡ã¯å»çãšãµãŒãã¹ãæ¹åããããã«ããŒã¿ ãµã€ãšã³ã¹ã®äœ¿çšãéå§ããŸãããç
æ°ãæ©æã«äºæž¬ã§ããã°ãå€ãã®å©ç¹ãåŸãããŸãã ãããã£ãŠããã®ããŒã¿ ãµã€ãšã³ã¹ ãããžã§ã¯ãã®ã¢ã€ãã¢ã§ã¯ãPython ã䜿çšããŠããŒãã³ãœã³ç
ãæ€åºããæ¹æ³ãåŠã³ãŸãã ããã¯ãéåã«åœ±é¿ãåãŒããéãã硬çŽãåŒãèµ·ããäžæ¢ç¥çµç³»ã®ç¥çµå€æ§æ§é²è¡æ§çŸæ£ã§ãã ããã¯è³å
ã®ããŒããã³çæãã¥ãŒãã³ã«åœ±é¿ãåãŒããã€ã³ãã§ã¯æ¯å¹Ž 1 äžäººä»¥äžã眹æ£ããŠããŸãã
èšèªïŒ Python
ããŒã¿ã»ãã/ããã±ãŒãž: UCI ML ããŒãã³ãœã³ç ããŒã¿ã»ãã
äžçšåºŠã®è€éãã®ããŒã¿ ãµã€ãšã³ã¹ ãããžã§ã¯ã
4. é³å£°ææ èªè
ããŒã¿ãµã€ãšã³ã¹ãµã³ãã«ãããžã§ã¯ãã®å®å
šãªå®è£
ããã§ãã¯ããŠãã ãã-
次ã«ãããŸããŸãªã©ã€ãã©ãªã®äœ¿çšæ¹æ³ãåŠã³ãŸãããã ãã®ããŒã¿ ãµã€ãšã³ã¹ ãããžã§ã¯ãã§ã¯ãé³å£°èªèã« librosa ã䜿çšããŸãã SER ã¯ãé³å£°ãã人éã®ææ
ãæ
åç¶æ
ãèå¥ããããã»ã¹ã§ãã ç§ãã¡ã¯å£°ã®ããŒã³ãšãããã䜿ã£ãŠææ
ãè¡šçŸãããããSER ã¯éèŠã§ãã ããããææ
ã¯äž»èŠ³çãªãã®ã§ãããããé³å£°æ³šéã¯å°é£ãªäœæ¥ã§ãã mfccãchromaãmel é¢æ°ã䜿çšããææ
èªèã«ã¯ RAVDESS ããŒã¿ã»ããã䜿çšããŸãã ãã®ã¢ãã«ã® MLPC åé¡åšãäœæããŸãã
èšèªïŒ Python
ããŒã¿ã»ãã/ããã±ãŒãž: RAVDESS ããŒã¿ã»ãã
5. æ§å¥ãšå¹Žéœ¢ã®æ€åº
ææ°ã®ããŒã¿ ãµã€ãšã³ã¹ ãããžã§ã¯ãã§éçšäž»ã«å¥œå°è±¡ãäžãã -
ãã㯠Python ã䜿ã£ãèå³æ·±ãããŒã¿ ãµã€ãšã³ã¹ã§ãã ãã£ã XNUMX æã®ç»åã䜿çšããŠã人ã®æ§å¥ãšå¹Žéœ¢ãäºæž¬ããæ¹æ³ãåŠã³ãŸãã ããã§ã¯ãã³ã³ãã¥ãŒã¿ãŒ ããžã§ã³ãšãã®åçã«ã€ããŠçŽ¹ä»ããŸãã æ§ç¯ããŸã
èšèªïŒ Python
ããŒã¿ã»ãã/ããã±ãŒãž: ã¢ãã£ãšã³ã¹
6. ãŠãŒããŒã®ããŒã¿åæ
ããŒã¿ãµã€ãšã³ã¹ãããžã§ã¯ãã®å®å
šãªå®è£
ããœãŒã¹ã³ãŒããšãšãã«è¡šç€ºããŸã-
ãã㯠ggplot2 ã䜿çšããããŒã¿èŠèŠåãããžã§ã¯ãã§ãããR ãšãã®ã©ã€ãã©ãªã䜿çšããŠããŸããŸãªãã©ã¡ãŒã¿ãŒãåæããŸãã Uber Pickups New York ããŒã¿ã»ããã䜿çšãã幎éã®ããŸããŸãªæéæ ã®ããžã¥ã¢ã©ã€ãŒãŒã·ã§ã³ãäœæããŸãã ããã¯ãæéãã«ã¹ã¿ããŒãžã£ãŒããŒã«ã©ã®ãããªåœ±é¿ãäžãããã瀺ããŠããŸãã
èšèªïŒ R
ããŒã¿ã»ãã/ããã±ãŒãž: ãã¥ãŒãšãŒã¯åžã® Uber éè¿ããŒã¿ã»ãã
7.ãã©ã€ããŒã®ç æ°æ€ç¥
ãããããŒã¿ãµã€ãšã³ã¹ãããžã§ã¯ãã«åãçµãã§ã¹ãã«ãã¢ããã°ã¬ãŒã -
å±
ç ãé転ã¯éåžžã«å±éºã§ãå±
ç ãé転ã«ããäºæ
ã¯æ¯å¹ŽçŽXNUMX件çºçããŠããŸãã ãã® Python ãããžã§ã¯ãã§ã¯ãç ããã©ã€ããŒãæ€åºããããŒãé³ã§èŠåã§ããã·ã¹ãã ãäœæããŸãã
ãã®ãããžã§ã¯ã㯠Keras ãš OpenCV ã䜿çšããŠå®è£ ãããŠããŸãã OpenCV ã䜿çšããŠé¡ãšç®ãæ€åºããKeras ã®å©ããåããŠããã£ãŒã ãã¥ãŒã©ã« ãããã¯ãŒã¯ææ³ã䜿çšããŠç®ã®ç¶æ (éããŠãããéããŠããã) ãåé¡ããŸãã
8.ãã£ããããã
Python ã§ãã£ããããããæ§ç¯ãããã£ãªã¢ãäžæ©åé²ãããŸããã -
ãã£ãããããã¯ããžãã¹ã«äžå¯æ¬ ãªèŠçŽ ã§ãã å€ãã®äŒæ¥ã¯é¡§å®¢ã«ãµãŒãã¹ãæäŸããå¿
èŠãããã顧客ã«ãµãŒãã¹ãæäŸããã«ã¯å€ãã®äººçè³æºãæéãåŽåãããããŸãã ãã£ãããããã¯ã顧客ããã®ãããã質åã®ããã€ãã«çããããšã§ã顧客ãšã®ããåãã®å€ããèªååã§ããŸãã ãã£ãããããã«ã¯åºæ¬çã«ããã¡ã€ã³åºæãšãªãŒãã³ãã¡ã€ã³ã® XNUMX çš®é¡ããããŸãã ãã¡ã€ã³åºæã®ãã£ãããããã¯ãç¹å®ã®åé¡ã解決ããããã«ãã䜿çšãããŸãã ãããã£ãŠãçŸå Žã§å¹æçã«æ©èœããããã«ã«ã¹ã¿ãã€ãºããå¿
èŠããããŸãã ãªãŒãã³ãã¡ã€ã³ã®ãã£ãããããã«ã¯ãããã質åãã§ããããããã¬ãŒãã³ã°ã«ã¯å€§éã®ããŒã¿ãå¿
èŠã«ãªããŸãã
ããŒã¿ã»ãã: ã€ã³ãã³ãã®jsonãã¡ã€ã«
èšèªïŒ Python
å é²çãªããŒã¿ãµã€ãšã³ã¹ãããžã§ã¯ã
9. ç»åãã£ãã·ã§ã³ãžã§ãã¬ãŒã¿ãŒ
ãœãŒã¹ã³ãŒãã䜿çšããŠå®å
šãªãããžã§ã¯ãå®è£
ããã§ãã¯ããŠãã ãã-
ç»åã®å
容ã説æããã®ã¯äººéã«ãšã£ãŠã¯ç°¡åãªäœæ¥ã§ãããã³ã³ãã¥ãŒã¿ã«ãšã£ãŠç»åã¯åãã¯ã»ã«ã®è²ã®å€ãè¡šãæ°å€ã®éåã«ãããŸããã ããã¯ã³ã³ãã¥ãŒã¿ã«ãšã£ãŠã¯é£ããäœæ¥ã§ãã ç»åã®å
容ãç解ããèªç¶èšèªã«ãã説æ (è±èªãªã©) ãäœæããããšããŸãé£ããäœæ¥ã§ãã ãã®ãããžã§ã¯ãã§ã¯ãç³ã¿èŸŒã¿ãã¥ãŒã©ã« ãããã¯ãŒã¯ (CNN) ãšãªã«ã¬ã³ã ãã¥ãŒã©ã« ãããã¯ãŒã¯ (LSTM) ãå®è£
ãã深局åŠç¿æè¡ã䜿çšããŠãç»åèšè¿°ãžã§ãã¬ãŒã¿ãŒãäœæããŸãã
ããŒã¿ã»ãã: ããªãã«ãŒ8K
èšèªïŒ Python
ãã¬ãŒã ã¯ãŒã¯ïŒ ã±ã©ã¹
10. ã¯ã¬ãžããã«ãŒãäžæ£è¡çºã®æ€åº
ããŒã¿ãµã€ãšã³ã¹ãããžã§ã¯ãã®ã¢ã€ãã¢ã«åãçµãã§æåãå°œãããŠãã ãã-
ãããŸã§ã§ãã¡ãœãããšæŠå¿µãç解ãå§ããŸããã ããã€ãã®é«åºŠãªããŒã¿ ãµã€ãšã³ã¹ ãããžã§ã¯ãã«ç§»ããŸãããã ãã®ãããžã§ã¯ãã§ã¯ã次ã®ãããªã¢ã«ãŽãªãºã ãåãã R èšèªã䜿çšããŸãã
èšèªïŒ R
ããŒã¿ã»ãã/ããã±ãŒãž: ã«ãŒãååŒããŒã¿ã»ãã
11. æ ç»æšèŠã·ã¹ãã
ãœãŒã¹ ã³ãŒãã䜿çšããŠæé«ã®ããŒã¿ ãµã€ãšã³ã¹ ãããžã§ã¯ãã®å®è£
ãæ¢çŽ¢ããŸã -
ãã®ããŒã¿ ãµã€ãšã³ã¹ ãããžã§ã¯ãã§ã¯ãR ã䜿çšããŠãæ©æ¢°åŠç¿ãéããŠæ ç»ã®ã¬ã³ã¡ã³ããŒã·ã§ã³ãå®è¡ããŸãã ã¬ã³ã¡ã³ããŒã·ã§ã³ ã·ã¹ãã ã¯ãä»ã®ãŠãŒã¶ãŒã®å¥œã¿ãé²èŠ§å±¥æŽã«åºã¥ããã£ã«ã¿ãªã³ã° ããã»ã¹ãéããŠãŠãŒã¶ãŒã«ææ¡ãéä¿¡ããŸãã A ãš B ãããŒã ã¢ããŒã³ã奜ãã§ãB ãããŒã³ ã¬ãŒã«ãºã奜ããªããA ãææ¡ã§ããŸãã圌ãããããæ°ã«å
¥ããããããŸããã ããã«ãããã¯ã©ã€ã¢ã³ãã¯ãã©ãããã©ãŒã ãšå¯Ÿè©±ã§ããããã«ãªããŸãã
èšèªïŒ R
ããŒã¿ã»ãã/ããã±ãŒãž: MovieLensããŒã¿ã»ãã
12. 顧客ã®ã»ã°ã¡ã³ããŒã·ã§ã³
ããŒã¿ ãµã€ãšã³ã¹ ãããžã§ã¯ã (ãœãŒã¹ ã³ãŒããå«ã) ã§éçšäž»ã«å¥œå°è±¡ãäžãã -
賌å
¥è
ã®ã»ã°ã¡ã³ããŒã·ã§ã³ã¯äººæ°ã®ããã¢ããªã±ãŒã·ã§ã³ã§ã
èšèªïŒ R
ããŒã¿ã»ãã/ããã±ãŒãž: Mall_Customers ããŒã¿ã»ãã
13. ä¹³ããã®åé¡
Python ã§ã®ããŒã¿ ãµã€ãšã³ã¹ ãããžã§ã¯ãã®å®å
šãªå®è£
ãã芧ãã ãã-
ããŒã¿ãµã€ãšã³ã¹ã®å»åŠçè²¢ç®ã«æ»ããPython ã§ä¹³ãããæ€åºããæ¹æ³ãåŠã³ãŸãããã IDC_normal ããŒã¿ã»ããã䜿çšããŠãä¹³ããã®æãäžè¬çãªåœ¢æ
ã§ãã浞最æ§ä¹³ç®¡ãããæ€åºããŸãã 乳管å
ã§çºçãã乳管ã®å€åŽã®ä¹³è
ºã®ç·ç¶çµç¹ãŸãã¯èèªçµç¹ã«æµžéããŸãã ãã®ããŒã¿åéç§åŠãããžã§ã¯ãã®ã¢ã€ãã¢ã§ã¯ã次ã®ãã®ã䜿çšããŸãã
èšèªïŒ Python
ããŒã¿ã»ãã/ããã±ãŒãž: IDC_ã¬ã®ã¥ã©ãŒ
14. 亀éæšèã®èªè
ããŒã¿ãµã€ãšã³ã¹ãããžã§ã¯ãã§èªåé転è»æè¡ã®ç²ŸåºŠãå®çŸ
éè·¯æšèãšäº€éã«ãŒã«ã¯ããã¹ãŠã®ãã©ã€ããŒã«ãšã£ãŠäºæ
ãé¿ããããã«éåžžã«éèŠã§ãã ã«ãŒã«ã«åŸãã«ã¯ããŸãéè·¯æšèãã©ã®ãããªãã®ããç解ããå¿
èŠããããŸãã 人ã¯ãè»äž¡ãé転ããæš©å©ãåŸãåã«ããã¹ãŠã®éè·¯æšèãåŠç¿ããå¿
èŠããããŸãã ãããçŸåšãèªåé転è»ã®æ°ã¯å¢å ããŠãããè¿ãå°æ¥ã人ã¯èªåã§è»ãé転ã§ããªããªããŸãã éè·¯æšèèªèãããžã§ã¯ãã§ã¯ãããã°ã©ã ãç»åãå
¥åãšããŠååŸããŠéè·¯æšèã®çš®é¡ãèªèããæ¹æ³ãåŠã³ãŸãã German Road Sign Recognition Reference Dataset (GTSRB) ã¯ã亀éæšèãå±ããã¯ã©ã¹ãèªèããããã®ãã£ãŒã ãã¥ãŒã©ã« ãããã¯ãŒã¯ãæ§ç¯ããããã«äœ¿çšãããŸãã ã¢ããªã±ãŒã·ã§ã³ãšå¯Ÿè©±ããããã®åçŽãª GUI ãäœæããŠããŸãã
èšèªïŒ Python
ããŒã¿ã»ãã: GTRB (ãã€ãã®äº€éæšèèªèãã³ãããŒã¯)
ç¶ããèªã
ãã¬ãŒãã³ã° ãããžã§ã¯ãçšã® 52 åã®ããŒã¿ã»ãã ããã³ããšã³ãéå Ž: éçºè ã®ã¹ãã«ãèšç·Žãããããžã§ã¯ã (æ°èŠ 5 件 + æ§æ¥ 43 件) æãèå³æ·±ã IT ãã€ããã㯠ã€ã³ãã©ã°ã©ãã£ã㯠ããã 12
åºæïŒ habr.com