Table of Contents[Hūnā][Hōʻike]
- 1. Titanic
- 2. Hoʻokaʻawale Pua ʻIlani
- 3. Wehewehena Kumukuai o ka Hale o Boston
- 4. Ho'āʻo maikaʻi waina
- 5. Ka wanana makeke
- 6. Manaʻo Kiʻiʻoniʻoni
- 7. Hoʻouka ʻana i ka wānana eligibility
- 8. Nānā Manaʻo me ka hoʻohana ʻana i ka ʻikepili Twitter
- 9. Ka wānana kūʻai aku
- 10. ʻIke Nuhou Hoʻopunipuni
- 11. Manaʻo Kūʻai Kūʻai
- 12. Ka wānana Churn Customer
- 13. Ka wanana kuai ana o Wallmart
- 14. Ka Ikepili Uber
- 15. ʻIkepili Covid-19
- Panina
ʻO ke aʻo ʻana i ka mīkini kahi haʻawina maʻalahi o ka hoʻonaʻauao ʻana i kahi polokalamu kamepiula a i ʻole algorithm e hoʻomaikaʻi mālie i kahi hana kikoʻī i hōʻike ʻia ma kahi kiʻekiʻe. ʻO ka ʻike kiʻi, ka ʻike hoʻopunipuni, nā ʻōnaehana paipai, a me nā noi aʻo mīkini ʻē aʻe ua hōʻike ʻia he kaulana.
ʻO nā hana ML e maʻalahi a maʻalahi ka hana kanaka, mālama i ka manawa a me ka hōʻoia i ka hopena kiʻekiʻe. ʻO Google, ka ʻenekini huli kaulana loa o ka honua, hoʻohana aʻo aʻo.
Mai ka nānā ʻana i ka nīnau a ka mea hoʻohana a me ka hoʻololi ʻana i ka hopena ma muli o nā hopena i ka hōʻike ʻana i nā kumuhana a me nā hoʻolaha e pili ana i ka nīnau, aia nā ʻano koho like ʻole.
ʻAʻole mamao loa ka ʻenehana noʻonoʻo a hoʻoponopono iā ia iho i ka wā e hiki mai ana.
ʻO kekahi o nā ala maikaʻi loa e hoʻomaka ai, ʻo ia ke kiʻi lima a hoʻolālā i kahi papahana. No laila, ua hōʻuluʻulu mākou i kahi papa inoa o 15 mau papahana aʻo mīkini kiʻekiʻe no ka poʻe hoʻomaka e hoʻomaka iā ʻoe.
1. Titanic
Manaʻo pinepine ʻia ʻo ia kekahi o nā hana nui a hauʻoli loa no ka poʻe makemake e aʻo hou e pili ana i ke aʻo ʻana i ka mīkini. ʻO ka Titanic Challenge kahi papahana aʻo mīkini kaulana a he ala maikaʻi hoʻi ia e ʻike ai i ka paepae ʻepekema data Kaggle. Hoʻokumu ʻia ka ʻikepili Titanic me ka ʻikepili maoli mai ka hili ʻana o ka moku pōʻino.
Loaʻa nā kikoʻī e like me ka makahiki o ke kanaka, ke kūlana socioeconomic, ke kāne, ka helu hale, ke awa haʻalele, a ʻo ka mea nui loa, inā i ola lākou!
Ua paʻa ka ʻenehana K-Nearest Neighbor a me ka mea hoʻoholo kumu lāʻau i nā hopena maikaʻi loa no kēia papahana. Inā ʻoe e ʻimi nei i kahi hoʻokūkū wikiwiki wikiwiki e hoʻomaikaʻi i kāu Hiki ke aʻo mīkini, nou keia ma Kaggle.
2. Hoʻokaʻawale Pua ʻIlani
Makemake nā poʻe hoʻomaka i ka papahana hoʻokaʻawale pua iris, a he wahi maikaʻi ia e hoʻomaka ai inā he mea hou ʻoe i ka aʻo mīkini. ʻO ka lōʻihi o nā sepals a me nā petals e hoʻokaʻawale i nā pua iris mai nā ʻano ʻē aʻe. ʻO ka manaʻo o kēia pāhana e hoʻokaʻawale i nā pua i ʻekolu ʻano: Virginia, setosa, a me Versicolor.
No ka hoʻomaʻamaʻa hoʻokaʻawale ʻana, hoʻohana ka papahana i ka waihona pua Iris, e kōkua i nā haumāna i ke aʻo ʻana i nā kumu o ka pili ʻana i nā waiwai helu a me nā ʻikepili. ʻO ka ʻikepili pua iris kahi mea liʻiliʻi hiki ke mālama ʻia i ka hoʻomanaʻo me ka ʻole o ka hoʻonui ʻana.
3. Huahelu Kumukuai o ka Hale o Bosetona
ʻO kekahi kaulana ʻikepili no nā mea hou i ke aʻo ʻana i ka mīkini ʻo ia ka ʻikepili hale noho ʻo Boston. ʻO kāna pahuhopu, ʻo ia ka wānana ʻana i nā waiwai home ma nā wahi like ʻole o Boston. Loaʻa ia i nā helu koʻikoʻi e like me ka makahiki, ka uku ʻauhau waiwai, ka nui o ka hewa, a me ka pili kokoke i nā kikowaena hana, hiki ke hoʻopilikia i nā kumu kūʻai hale.
He mea maʻalahi a liʻiliʻi ka ʻikepili, e maʻalahi ke hoʻokolohua me nā mea hou. No ka ʻike ʻana i nā mea e hoʻohuli ai i ke kumukūʻai waiwai ma Boston, hoʻohana nui ʻia nā ʻenehana regression ma nā ʻāpana like ʻole. He wahi maikaʻi ia e hoʻomaʻamaʻa i nā ʻenehana regression a loiloi i ka maikaʻi o kā lākou hana.
4. Hoao Waina
ʻO ka waina kahi mea inu ʻona ʻokoʻa e pono ai i nā makahiki o ka fermenting. ʻO ka hopena,ʻo ka'ōmole waina kahiko he waina kūʻai a kiʻekiʻe. ʻO ke koho ʻana i ka hue waina maikaʻi e pono ai nā makahiki o ka ʻike ʻana i ka waina, a hiki ke hana i kahi kaʻina hit-a-miss.
ʻO ka papahana hoʻāʻo maikaʻi waina e loiloi i nā waina me ka hoʻohana ʻana i nā hoʻokolohua physicochemical e like me ke kiʻekiʻe o ka waiʻona, ka acidity paʻa, density, pH, a me nā mea ʻē aʻe. Hoʻoholo ka papahana i nā pae a me ka nui o ka waina. ʻO ka hopena, lilo ke kūʻai ʻana i ka waina i ka makani.
5. Makeke Waiwai
He mea hoihoi kēia hana inā ʻoe e hana a i ʻole ma ka ʻoihana kālā. Ua aʻo nui ʻia ka ʻikepili mākeke e ka poʻe hoʻonaʻauao, ʻoihana, a me ke kumu o ka loaʻa kālā lua. He mea koʻikoʻi nō hoʻi ka hiki i ka ʻepekema ʻikepili ke aʻo a ʻimi i ka ʻikepili moʻo manawa. He wahi maikaʻi loa ka ʻikepili mai ka mākeke waihona.
ʻO ke kumu o ka hoʻoikaika ʻana ʻo ia ka wānana i ka waiwai o ka wā e hiki mai ana. Hoʻokumu ʻia kēia i ka hana mākeke o kēia manawa a me nā helu mai nā makahiki i hala. Ua hōʻiliʻili ʻo Kaggle i ka ʻikepili ma ka papa kuhikuhi NIFTY-50 mai ka makahiki 2000, a ke hōʻano hou ʻia nei i kēlā me kēia pule. Mai Ianuali 1, 2000, ua loaʻa nā kumukūʻai kumukūʻai no nā hui he 50.
6. Manaʻo Kiʻiʻoniʻoni
Manaʻo wau ua loaʻa iā ʻoe kēlā manaʻo ma hope o ka ʻike ʻana i kahi kiʻiʻoniʻoni maikaʻi. Ua manaʻo paha ʻoe i ka manaʻo e hoʻokalakupua i kou naʻau ma ka nānā ʻana i nā kiʻiʻoniʻoni like?
ʻIke mākou ua hoʻomaikaʻi nui nā lawelawe OTT e like me Netflix i kā lākou ʻōnaehana ʻōlelo. Ma ke ʻano he haumāna aʻo mīkini, pono ʻoe e hoʻomaopopo i ke ʻano o ka huli ʻana o ia mau algorithm i nā mea kūʻai aku ma muli o kā lākou makemake a me nā loiloi.
ʻO ka ʻikepili IMDB i hoʻonohonoho ʻia ma Kaggle ʻo ia paha kekahi o ka piha loa, e ʻae ana i nā hiʻohiʻona manaʻo e hoʻohālikelike ʻia ma muli o ke poʻo kiʻiʻoniʻoni, ka helu mea kūʻai aku, ke ʻano, a me nā mea ʻē aʻe. He ala maikaʻi loa ia e aʻo ai e pili ana i ka Content-Based Filtering and Feature Engineering.
7. Wehewehe ʻana i ke kūpono
Hoʻopili ka honua i nā hōʻaiʻē. ʻO ke kumu waiwai nui o nā panakō mai ka ukupanee ma nā hōʻaiʻē. No laila ʻo lākou kā lākou ʻoihana kumu.
Hiki i nā kānaka a i ʻole nā hui o nā kānaka ke hoʻonui i ka ʻoihana ma o ka hoʻokomo kālā ʻana i kahi paʻa me ka manaʻolana e ʻike i ka piʻi ʻana o ka waiwai i ka wā e hiki mai ana. He mea nui i kekahi manawa ke ʻimi i kahi hōʻaiʻē e hiki ai ke lawe i nā pilikia o kēia ʻano a komo pū i kekahi mau leʻaleʻa honua.
Ma mua o ka ʻae ʻia ʻana o kahi hōʻaiʻē, ʻoi aku ka paʻakikī o nā panakō. No ka mea he ʻano koʻikoʻi loa nā hōʻaiʻē i ke ola o ka poʻe he nui, ʻo ka wānana ʻana i ke kūpono no kahi hōʻaiʻē a kekahi e noi ai he mea maikaʻi loa ia, e ʻae ai i ka hoʻolālā maikaʻi ʻana ma mua o ka ʻae ʻia a hōʻole ʻia paha.
8. Nānā Manaʻo e hoʻohana ana i ka ʻikepili Twitter
mahalo aku i pūnaewele pūnaewele e like me Twitter, Facebook, a me Reddit, ua maʻalahi nā manaʻo extrapolating a me nā ʻano. Hoʻohana ʻia kēia ʻike e hoʻopau i nā manaʻo e pili ana i nā hanana, nā kānaka, nā haʻuki, a me nā kumuhana ʻē aʻe. Hoʻohana ʻia nā manaʻo hoʻonaʻauao mīkini pili i ka manaʻo i nā ʻano hoʻonohonoho like ʻole, me nā hoʻolaha politika a me nā loiloi huahana Amazon.
E nānā maikaʻi kēia papahana i kāu kōpili! No ka ʻike ʻana i ka naʻau a me ka nānā ʻana i nā hiʻohiʻona, hiki ke hoʻohana nui ʻia nā ʻenehana e like me ke kākoʻo ʻana i nā mīkini vector, regression, a me nā algorithms classification (ʻike i nā ʻoiaʻiʻo a me nā manaʻo).
9. Ka wanana kuai
Makemake nā ʻoihana B2C nui a me nā mea kālepa e ʻike i ka nui o kēlā me kēia huahana i kā lākou waihona. Kōkua ka wānana kūʻai i ka poʻe nona nā ʻoihana e hoʻoholo i nā mea i makemake nui ʻia. ʻO ka wānana kūʻai pololei e hoʻemi nui i ka hoʻopau ʻana aʻo ka hoʻoholo ʻana i ka hopena hoʻonui i nā waihona kālā e hiki mai ana.
Hoʻohana nā mea kūʻai kūʻai e like me Walmart, IKEA, Big Basket, a me Big Bazaar i ka wānana kūʻai e koho i ka makemake o ka huahana. Pono ʻoe e kamaʻāina i nā ʻenehana like ʻole o ka hoʻomaʻemaʻe ʻana i ka ʻikepili maka no ke kūkulu ʻana i ia mau papahana ML. Eia kekahi, pono ka ʻike maikaʻi o ka loiloi regression, ʻoi aku ka maʻalahi o ka regression linear.
No kēia mau ʻano hana, pono ʻoe e hoʻohana i nā hale waihona puke e like me Dora, Scrubadub, Pandas, NumPy, a me nā mea ʻē aʻe.
10. ʻIke Nuhou Hoʻopunipuni
ʻO ia kekahi hana aʻo mīkini ʻokiʻoki e pili ana i nā keiki kula. Ke laha nei ka lono hoʻopunipuni e like me ke ahi ahi, e like me kā mākou i ʻike ai. Loaʻa nā mea a pau ma ka pāpaho pūnaewele, mai ka hoʻopili ʻana i nā poʻe a hiki i ka heluhelu ʻana i ka nūhou i kēlā me kēia lā.
ʻO ka hopena, ua paʻakikī loa ka ʻike ʻana i nā lono hoʻopunipuni i kēia mau lā. Nui nā ʻupena pūnaewele nui, e like me Facebook a me Twitter, ua loaʻa nā algorithms i kahi e ʻike ai i nā nūhou hoʻopunipuni ma nā hoʻouna a me nā hānai.
No ka ʻike ʻana i nā nūhou wahaheʻe, pono kēia ʻano papahana ML i ka ʻike hohonu o nā ala NLP he nui a me nā algorithms classification (PassiveAggressiveClassifier a i ʻole Naive Bayes classifier).
11. Kuhikuhi Kūʻai Kūʻai
Ke noʻonoʻo nui nei nā mea kūʻai aku i ke kūʻai ʻana ma ka pūnaewele i ka wā i hoʻouka ai ka coronavirus i ka honua i 2020. ʻO ka hopena, ua koi ʻia nā hale kūʻai e hoʻololi i kā lākou ʻoihana ma ka pūnaewele.
ʻO nā mea kūʻai aku, ma ka ʻaoʻao ʻē aʻe, ke ʻimi nei lākou i nā makana nui, e like me lākou i loko o nā hale kūʻai, a ke ʻimi nui nei lākou i nā coupon super-saving. Aia kekahi mau pūnaewele i hoʻolaʻa ʻia no ka hana ʻana i nā coupons no kēlā mau mea kūʻai aku. Hiki iā ʻoe ke aʻo e pili ana i ka ʻimi ʻikepili ma ke aʻo ʻana i ka mīkini, ka hana ʻana i nā kiʻi paʻi, nā pakuhi pai, a me nā histograms e ʻike ai i ka ʻikepili, a me ka hana ʻenekinia me kēia papahana.
No ka hoʻoulu ʻana i nā wānana, hiki iā ʻoe ke nānā i nā ala imputation data no ka hoʻokele ʻana i nā waiwai NA a me ka like cosine o nā mea hoʻololi.
12. Manaʻo Kūkākūkā Churn
ʻO nā mea kūʻai aku ka waiwai nui o ka ʻoihana, a he mea koʻikoʻi ka mālama ʻana iā lākou no kēlā me kēia ʻoihana e manaʻo ana e hoʻonui i ka loaʻa kālā a kūkulu i nā pilina koʻikoʻi me lākou.
Eia kekahi, ʻoi aku ka nui o ke kumukūʻai no ka loaʻa ʻana o kahi mea kūʻai hou aku ma mua o ke kumu kūʻai no ka mālama ʻana i kahi mea kūʻai aku. ʻO ka Customer Churn/Attrition kahi pilikia ʻoihana kaulana kahi i hoʻōki ai nā mea kūʻai a i ʻole nā mea kākau inoa i ka hana ʻoihana me kahi lawelawe a i ʻole kahi hui.
ʻAʻole lākou e lilo hou i mea kūʻai uku. Manaʻo ʻia ka mea kūʻai aku i ka pōʻino inā he manawa kūikawā ia mai ka wā i launa pū ai ka mea kūʻai aku me ka hui. ʻO ka ʻike ʻana inā e churn ka mea kūʻai aku, a me ka hāʻawi wikiwiki ʻana i ka ʻike kūpono e pili ana i ka mālama ʻana i ka mea kūʻai aku, he mea koʻikoʻi ia e hoʻohaʻahaʻa i ka churn.
ʻAʻole hiki i ko mākou lolo ke kali i ka huli ʻana o ka mea kūʻai aku no nā miliona o nā mea kūʻai aku; eia kahi e kōkua ai ke aʻo ʻana i ka mīkini.
13. Ka wanana kuai ana o Wallmart
ʻO kekahi o nā noi koʻikoʻi o ke aʻo ʻana i ka mīkini ʻo ia ka wānana kūʻai, e pili ana i ka ʻike ʻana i nā hiʻohiʻona e hoʻohuli i ke kūʻai aku o nā huahana a me ka manaʻo ʻana i ka nui o nā kūʻai aku e hiki mai ana.
Hoʻohana ʻia ka ʻikepili Walmart, aia i nā ʻikepili kūʻai mai 45 mau wahi, i kēia haʻawina aʻo mīkini. Hoʻokomo ʻia nā kūʻai aku ma kēlā me kēia hale kūʻai, ma ka ʻāpana, i kēlā me kēia pule i ka waihona. ʻO ke kumu o kēia papahana aʻo mīkini e kali i nā kūʻai aku no kēlā me kēia keʻena i kēlā me kēia puka i hiki iā lākou ke hana i nā hoʻoholo hoʻoholo hoʻolālā hoʻolālā waiwai.
He paʻakikī ka hana ʻana me ka ʻikepili Walmart no ka mea aia nā hanana markdown i koho ʻia e pili ana i ke kūʻai aku a pono e noʻonoʻo ʻia.
14. ʻIkepili ʻikepili Uber
I ka wā e pili ana i ka hoʻokō ʻana a me ka hoʻohui ʻana i ka aʻo mīkini a me ke aʻo hohonu i kā lākou mau polokalamu, ʻaʻole mamao loa ka lawelawe kaʻana kaʻa kaulana. I kēlā me kēia makahiki, hana ʻo ia i nā piliona huakaʻi, e ʻae ana i ka poʻe commuters e hele i kēlā me kēia manawa o ke ao a me ka pō.
No ka mea he waihona mea kūʻai nui kona, pono ia i ka lawelawe mea kūʻai aku e hoʻoponopono i nā hoʻopiʻi o nā mea kūʻai aku me ka wikiwiki.
Loaʻa iā Uber kahi waihona o nā miliona o nā koho koho e hiki iā ia ke hoʻohana no ka nānā ʻana a hōʻike i nā huakaʻi mea kūʻai aku e wehe i nā ʻike a hoʻomaikaʻi i ka ʻike mea kūʻai aku.
15. Nānā Covid-19
Ua kāhili ʻo COVID-19 i ka honua i kēia lā, ʻaʻole wale ma ke ʻano o kahi maʻi maʻi. ʻOiai ke kau nei ka poʻe loea lapaʻau i ka hana ʻana i nā lāʻau lapaʻau kūpono a me ka pale ʻana i ka honua, ʻepekema ʻikepili ʻaʻole mamao hope.
Hoʻolaha ʻia nā hihia hou, ka helu hana i kēlā me kēia lā, nā mea make, a me nā helu hoʻāʻo. Hana ʻia nā wānana i kēlā me kēia lā e pili ana i ka puka ʻana o SARS o ke kenekulia i hala. No kēia, hiki iā ʻoe ke hoʻohana i ka loiloi regression a kākoʻo i nā hiʻohiʻona wānana e pili ana i ka mīkini vector.
Panina
No ka hōʻuluʻulu ʻana, ua kūkākūkā mākou i kekahi o nā papahana ML kiʻekiʻe e kōkua iā ʻoe i ka hoʻāʻo ʻana i ka papahana Machine Learning a me ka hopu ʻana i kāna mau manaʻo a me ka hoʻokō. ʻO ka ʻike pehea e hoʻohui ai i ka Machine Learning hiki ke kōkua iā ʻoe e holomua i kāu ʻoihana e like me ka lawe ʻana o ka ʻenehana i kēlā me kēia ʻoihana.
ʻOiai e aʻo ana i ka Machine Learning, paipai mākou iā ʻoe e hoʻomaʻamaʻa i kāu mau manaʻo a kākau i kāu mau algorithm āpau. ʻOi aku ka mea nui o ke kākau ʻana i nā algorithm aʻo ke aʻo ʻana ma mua o ka hana ʻana i kahi papahana, a hāʻawi pū kekahi iā ʻoe i kahi pōmaikaʻi i ka hoʻomaopopo pono ʻana i nā kumuhana.
Waiho i ka Reply