Aloha, Habr! Ke hāʻawi aku nei au iā ʻoe i kahi unuhi o ka ʻatikala na kristencarter7519.
ʻOiai me he mea lā he mau lā wale nō mākou mai 2020, he mea koʻikoʻi kēia mau lā ma ke kahua o ka hoʻomohala polokalamu. Ma kēia ʻatikala, e ʻike mākou pehea e hoʻololi ai ka makahiki 2020 i ke ola o nā mea hoʻomohala polokalamu.
Aia ka wā e hiki mai ana o ka hoʻomohala polokalamu!
ʻO ka hoʻomohala polokalamu maʻamau ka hoʻomohala ʻana i nā polokalamu ma ke kākau ʻana i nā code ma muli o kekahi mau lula paʻa. Akā ua ʻike ka hoʻomohala ʻana i nā polokalamu hou i ka neʻe ʻana o ka paradigm me ka holomua o ka naʻauao artificial, aʻo mīkini a me ke aʻo hohonu. Ma ka hoʻohui ʻana i kēia mau ʻenehana ʻekolu, hiki i nā mea hoʻomohala ke hana i nā ʻōnaehana polokalamu e aʻo mai nā ʻōlelo aʻo a hoʻohui i nā hiʻohiʻona a me nā hiʻohiʻona hou i ka ʻikepili e pono ai e hana i ka hopena i makemake ʻia.
E ho'āʻo kākou me kekahi code
I ka wā lōʻihi, ua lilo nā ʻōnaehana hoʻomohala pūnaewele neural i mea paʻakikī e pili ana i ka hoʻohui ʻana a me nā pae o ka hana a me nā pilina. Hiki i nā mea hoʻomohala, no ka laʻana, ke kūkulu i kahi pūnaewele neural maʻalahi me Python 3.6. Eia kekahi papahana laʻana e hana ana i ka hoʻohālikelike binary me 1s a i ʻole 0s.
ʻOiaʻiʻo, hiki iā mākou ke hoʻomaka ma ka hana ʻana i kahi papa neural network:
lawe mai iā NumPy ma ke ʻano he NP
X=np.array([[0,1,1,0],[0,1,1,1],[1,0,0,1]])
y=np.array([[0],[1],[1]])Hoʻohana i ka hana sigmoid:
def sigmoid ():
return 1/(1 + np.exp(-x))
def derivatives_sigmoid ():
return x * (1-x)Hoʻomaʻamaʻa ʻana i kahi kumu hoʻohālike me nā kaupaona mua a me nā ʻano like ʻole:
epoch=10000
lr=0.1
inputlayer_neurons = X.shape[1]
hiddenlayer_neurons = 3
output_neurons = 1
wh=np.random.uniform(size=(inputlayer_neurons,hiddenlayer_neurons))
bh=np.random.uniform(size=(1,hiddenlayer_neurons))
wout=np.random.uniform(size=(hiddenlayer_neurons,output_neurons))
bout=np.random.uniform(size=(1,output_neurons))No ka poʻe hoʻomaka, inā makemake ʻoe i ke kōkua e pili ana i nā pūnaewele neural, hiki iā ʻoe ke ʻimi i ka pūnaewele no nā pūnaewele o nā ʻoihana hoʻomohala polokalamu kiʻekiʻe a i ʻole hiki iā ʻoe ke hoʻolimalima i nā mea hoʻomohala AI/ML e hana i kāu papahana.
Hoʻololi code me ka hoʻohana ʻana i kahi neuron papa puka
hidden_layer_input1=np.dot(X,wh)
hidden_layer_input=hidden_layer_input1 + bh
hiddenlayer_activations = sigmoid(hidden_layer_input)
output_layer_input1=np.dot(hiddenlayer_activations,wout)
output_layer_input= output_layer_input1+ bout
output = sigmoid(output_layer_input)
Ua hewa ka helu ʻana no ka papa helu huna
E = y-output
slope_output_layer = derivatives_sigmoid(output)
slope_hidden_layer = derivatives_sigmoid(hiddenlayer_activations)
d_output = E * slope_output_layer
Error_at_hidden_layer = d_output.dot(wout.T)
d_hiddenlayer = Error_at_hidden_layer * slope_hidden_layer
wout += hiddenlayer_activations.T.dot(d_output) *lr
bout += np.sum(d_output, axis=0,keepdims=True) *lr
wh += X.T.dot(d_hiddenlayer) *lr
bh += np.sum(d_hiddenlayer, axis=0,keepdims=True) *lrHōʻalo
print (output)
[[0.03391414]
[0.97065091]
[0.9895072 ]]He mea pono ke hoʻomau i ka manawa me nā ʻōlelo hoʻonohonoho hou a me nā ʻenehana coding, a pono e ʻike nā mea papahana i nā mea hana hou e kōkua ai i kā lākou mau polokalamu e pili ana i nā mea hoʻohana hou.
I ka makahiki 2020, pono e noʻonoʻo nā mea hoʻomohala polokalamu e hoʻokomo i kēia mau mea hana hoʻomohala polokalamu 5 i kā lākou huahana, ʻaʻohe mea e pili ana i ka ʻōlelo papahana a lākou e hoʻohana ai:
1. Hoʻoponopono ʻŌlelo Kūlohelohe (NLP)
Me kahi chatbot e hoʻoheheʻe i ka lawelawe o ka mea kūʻai aku, loaʻa iā NLP ka manaʻo o nā polokalamu polokalamu e hana ana i ka hoʻomohala polokalamu hou. Hoʻohana lākou i nā mea hana NLTK e like me Python NLTK e hoʻokomo koke i ka NLP i loko o nā chatbots, nā mea kōkua kikohoʻe, a me nā huahana kikohoʻe. Ma ka waena o 2020 a i ka wā e hiki mai ana, e ʻike ʻoe i ka lilo ʻana o NLP i mea nui i nā mea āpau mai nā ʻoihana kūʻai aku i nā kaʻa a me nā mea hana no ka home a me ke keʻena.
Ke neʻe nei i mua me nā mea hana a me nā ʻenehana hoʻomohala ʻoi aku ka maikaʻi, hiki iā ʻoe ke manaʻo e hoʻohana nā mea hoʻomohala polokalamu iā NLP ma nā ʻano like ʻole, mai nā mea hoʻohana leo-leo a hiki i ka hoʻokele menu maʻalahi, ka nānā ʻana i ka manaʻo, ka ʻike pōʻaiapili, ka manaʻo, a me ka ʻike ʻike. E loaʻa kēia mau mea āpau i ka hapa nui o nā mea hoʻohana, a hiki i nā hui ke hoʻokō i ka ulu ʻana o ka huahana a hiki i $ 430 biliona e 2020 (e like me ka IDC, i haʻi ʻia e Deloitte).
2. GraphQL e pani ana ia REST Apis
Wahi a nā mea hoʻomohala ma kaʻu paʻa, ʻo ia kahi hui hoʻomohala polokalamu ma waho, ke nalowale nei ka REST API i kona mana ma luna o ka honua noi ma muli o ka lohi o ka hoʻouka ʻana i nā ʻikepili e pono e hana ʻia mai nā URL he nui i kēlā me kēia.
He ʻano hou ʻo GraphQL a he ʻokoʻa ʻoi aku ka maikaʻi i ka hale hoʻolālā REST-based e kiʻi i nā ʻikepili pili āpau mai nā pūnaewele lehulehu me ka hoʻohana ʻana i hoʻokahi nīnau. Hoʻomaikaʻi kēia i ka launa pū ʻana o ka mea kūʻai aku a hoʻemi i ka latency, e ʻoi aku ka pane o ka noi no ka mea hoʻohana.
Hiki iā ʻoe ke hoʻomaikaʻi i kāu mākau hoʻomohala polokalamu ke hoʻohana ʻoe iā GraphQL no ka hoʻomohala polokalamu. Eia hou, pono ia i ka code liʻiliʻi ma mua o REST Api a hiki iā ʻoe ke hana i nā nīnau paʻakikī i kekahi mau laina maʻalahi. Hiki ke hoʻolako pū ʻia me kekahi mau hiʻohiʻona Backand as a Service (BaaS) e maʻalahi ke hoʻohana ʻia e nā mea hoʻomohala polokalamu ma nā ʻōlelo hoʻonohonoho like ʻole, me Python, Node.js, C++ a me Java.
3. Haʻahaʻa coding pae/ʻaʻohe code (haʻahaʻa code)
Hāʻawi nā mea hana hoʻomohala polokalamu haʻahaʻa haʻahaʻa i nā pono he nui. Pono ia e like me ka hiki ke kākau i nā papahana he nui mai ka wā kahiko. Hāʻawi ka code haʻahaʻa i ka code preconfigured i hiki ke hoʻokomo ʻia i loko o nā papahana nui. ʻAe kēia i nā mea polokalamu ʻole e hana wikiwiki a maʻalahi i nā huahana paʻakikī a hoʻolalelale i ka ʻōnaehana hoʻomohala hou.
Wahi a kahi hōʻike TechRepublic, ua hoʻohana ʻia nā mea hana no-code/low code i nā puka pūnaewele, nā ʻōnaehana polokalamu, nā noi kelepona a me nā wahi ʻē aʻe. E ulu ka mākeke hāmeʻa haʻahaʻa i $ 15 biliona e 2020. Hoʻohana kēia mau mea hana i nā mea āpau, me ka hoʻokele ʻana i ka loiloi workflow, kānana ʻikepili, lawe mai a hoʻokuʻu aku. Eia nā papa helu haʻahaʻa haʻahaʻa maikaʻi loa ma 2020:
- Microsoft PowerApps
- Mendiks
- Nā ʻōnaehana waho
- Mea hana Zoho
- Kapua App Salesforce
- Base wikiwiki
- ʻ bootpala puna
4. 5G nalu
ʻO ka pilina 5G e hopena nui i ka polokalamu kelepona a me ka hoʻomohala polokalamu a me ka hoʻomohala pūnaewele. Ma hope o nā mea a pau, me nā ʻenehana e like me IoT, pili nā mea āpau. No laila, e hoʻohana ka polokalamu kelepona i ka hapa nui o nā hiki o nā pūnaewele uea kiʻekiʻe me 5G.
Ma kahi ninaninau hou me Digital Trends, ua ʻōlelo ʻo Dan Dery, ka hope pelekikena o nā huahana o Motorola, "i nā makahiki e hiki mai ana, e hāʻawi ʻo 5G i ka ʻikepili wikiwiki, ka bandwidth kiʻekiʻe, a e hoʻolalelale i nā polokalamu kelepona 10 mau manawa ʻoi aku ka wikiwiki ma mua o nā ʻenehana uila i loaʻa."
Ma kēia mālamalama, e hana nā ʻoihana polokalamu e lawe i ka 5G i nā noi hou. I kēia manawa, ʻoi aku ma mua o 20 mau mea hana i hoʻolaha i nā hoʻonui i kā lākou pūnaewele. No laila, e hoʻomaka ana nā mea hoʻomohala e hana i ka hoʻohana ʻana i nā API kūpono e hoʻohana pono i ka 5G. E hoʻomaikaʻi nui ka ʻenehana i kēia mau mea:
- ʻO ka palekana o ka polokalamu pūnaewele, ʻoi loa no ka Network Slicing.
- Hāʻawi i nā ala hou e mālama ai i nā ID mea hoʻohana.
- Hāʻawi iā ʻoe e hoʻohui i nā hana hou i nā noi me ka latency haʻahaʻa.
- E hoʻoikaika i ka hoʻomohala ʻana o ka ʻōnaehana AR/VR.
5. ʻO ka hōʻoia maʻalahi
Ke lilo nei ka hōʻoia i kaʻina hana kūpono no ka pale ʻana i ka ʻikepili koʻikoʻi. ʻAʻole pilikia wale ka ʻenehana paʻakikī i nā hacks software, akā kākoʻo pū kekahi i ka naʻauao artificial a me ka computing quantum. Akā ke ʻike nei ka mākeke hoʻomohala polokalamu i nā ʻano hōʻoia hou, e like me ka loiloi leo, biometrics a me ka ʻike maka.
I kēia manawa, ʻimi nā mea hackers i nā ala like ʻole e hoʻopaʻa i nā ID mea hoʻohana pūnaewele a me nā ʻōlelo huna. No ka mea ua maʻa mua nā mea hoʻohana kelepona i ke komo ʻana i kā lākou kelepona me ka manamana lima a i ʻole ka nānā ʻana i ka maka, no laila e hoʻohana ana i nā mea hana hōʻoia, ʻaʻole lākou e pono i nā mana hōʻoia hou e like me ka liʻiliʻi o ka ʻaihue cyber. Eia kekahi mau mea hana hōʻoia multi-factor me SSL encryption.
- Hoʻololi ʻo Soft Tokens i kāu mau kelepona i mau mea hōʻoia multi-factor kūpono.
- He mea maʻalahi ka hoʻohana ʻana a me ke ʻano kaulana o nā mea hōʻoia ʻo EGrid i ka ʻoihana.
- ʻO kekahi o nā papahana hōʻoia maikaʻi loa no nā ʻoihana ʻo RSA SecurID Access, OAuth, Ping Identity, Authx, a me Aerobase.
Aia nā ʻoihana polokalamu ma India a me ʻAmelika e hana nei i ka noiʻi nui ma ke kahua o ka hōʻoia a me ka biometrics. Ke hoʻolaha nei lākou iā AI e hana i nā polokalamu ʻoi aku ka maikaʻi no ka leo, face-id, behavioral and biometric authentication. I kēia manawa hiki iā ʻoe ke pale i nā kaha kikohoʻe a hoʻomaikaʻi i ka hiki ke kahua.
hopena
Me he mea lā e liʻiliʻi ka paʻakikī o ke ola no nā polokalamu polokalamu ma 2020 no ka mea e wikiwiki ana ka wikiwiki o ka hoʻomohala polokalamu. E maʻalahi ka hoʻohana ʻana i nā mea hana i loaʻa. ʻO ka hope loa, e hana kēia holomua i kahi honua ikaika e komo ana i kahi makahiki kikohoʻe hou.
Source: www.habr.com