Unoda kutanga nazvo machine learning?
Ini ndakagadzira chidzidzo chakareruka uye chiri nyore kune vanotanga zvakakwana. Pamwe chete, tichaenda pamusoro pematanho ekutanga ekudzidzisa modhi yekudzidza muchina.
Ndichiri kutsanangura matanho ekudzidzisa modhi rimwe nerimwe, ini ndichapawo muenzaniso chaiwo wedambudziko rekudzidza muchina zvakare. Saka, kana iwe uchida kutevedzera, unogona kudhawunirodha iyi sampuli data seti kubva pane ino batanidzo.
Iyi ingori yemuenzaniso dataset yekukubatsira kuti utange nekudzidza muchina.
Tine 18 kukosha kwevanhu vemazera akasiyana uye varume nevakadzi vane mimhanzi yavanofarira inotsanangurwa. Nekushandisa, maficha e "zera" uye "murume" isu tichaedza kufungidzira kuti ndeupi rudzi rwemimhanzi ravanofarira.
Cherechedzo: 1 uye 0 anopihwa kune varume semukadzi nemurume mune ino dataset.
Nekudaro, kana iwe usingade kutevedzera muenzaniso, zvakare zvakanaka. Ndichange ndichitsanangura ese aya matanho zvakadzama. Saka, ngatinyure mukati!
Zvinhu Zvokutanga Kuziva
Tisati tapinda mumatanho ekudzidzisa modhi, ngatijekese mamwe mapoinzi. Kudzidza muchina chinhu chakagadzirwa njere chirango chinotarisa pakugadzira maalgorithms anogona kudzidza kubva kune data.
Kuti uite izvi, modhi yekudzidza yemuchina inodzidziswa pane dataset inodzidzisa modhi maitiro ekufanotaura chaiko kana kupatsanura pane nyowani, data yakambozivikanwa kare.
Saka, ndedzipi idzi modhi? A muchina wekudzidza modhi yakafanana neresipi inoshandiswa nekombuta kugadzira data yekufungidzira kana sarudzo.
Muenzaniso, seresipi, unotevera seti yemirairo yekuongorora data uye kugadzira fungidziro kana mitongo yakavakirwa pamapateni anowanikwa mune data. Iyo yakawanda data iyo modhi inodzidziswa pairi, iyo yakanyanya kujeka kufanotaura kwayo kunova.
Ndeupi Rudzi Rwatinogona Kudzidzisa?
Ngationei kuti ndedzipi mhando dzekutanga dzekudzidza muchina.
- Linear Regression: modhi inofanotaura chinoenderera chinangwa chinoshanduka kubva kune chimwe kana akawanda ekuisa akasiyana.
- Neural Networks: network yemanode akabatanidzwa anogona kudzidza kuona maitiro akaoma mu data.
- Miti Yechisarudzo: maitiro ekuita sarudzo akavakirwa paketani yebazi kana-zvimwe zvirevo.
- Clustering: seti yemamodheru anounganidza mapoinzi edata anoenzaniswa zvichienderana nekufanana.
- Logistic Regression: muenzaniso wematambudziko emhando yebhinari umo chinangwa chinoshanduka chine maitiro maviri anogona kuve.
- Miti Yechisarudzo: maitiro ekuita sarudzo akavakirwa paketani yebazi kana-zvimwe zvirevo.
- Random Sango: ensemble modhi inoumbwa nemiti yakawanda yesarudzo. Ivo vanowanzo shandiswa kupatsanura uye regression application.
- K-Vavakidzani Vepedyo: modhi inofanotaura chinangwa chinoshanduka uchishandisa k-padyo data mapoinzi museti yekudzidziswa.
Zvichienderana nedambudziko redu uye dataset, isu tinosarudza kuti ndeipi modhi yekudzidza yemuchina inoenderana nemamiriro edu zvakanyanya. Asi, tichazodzoka kune izvi gare gare. Zvino, ngatitange kudzidzisa modhi yedu. Ndinovimba wakatodhaunirodha dhatabheti kana uchida kutevera muenzaniso wedu.
Zvakare, ndinokurudzira kuve Jupyter Notebook yakaiswa pamushini wako wepanzvimbo uye uchiishandisa kumapurojekiti ako ekudzidza muchina.
1: Tsanangura dambudziko
Danho rekutanga mu kudzidzisa muchina kudzidza modhi iri kutsanangura nyaya inofanira kugadziriswa. Izvi zvinosanganisira kusarudza mavhezheni aunoda kufanotaura (anozivikanwa seanotariswa shanduko) uye akasiyana anozoshandiswa kugadzira iwo mafungidziro (anozivikanwa semaficha kana kufanotaura).
Iwe unofanirwawo kusarudza rudzi rwedambudziko rekudzidza-muchina rauri kuyedza kugadzirisa (kuronga, kudzoreredza, kubatanidza, zvichingodaro) uye rudzi rwe data rauchazoda kuunganidza kana kuwana kudzidzisa modhi yako.
Rudzi rwemodhi yaunoshandisa ichatemwa nerudzi rwemushini wekudzidza dambudziko rauri kuda kugadzirisa. Classification, regression, uye clustering ndiwo mapoka matatu ekutanga e Matambudziko ekudzidza muchina. Paunenge uchida kufanotaura shanduko yemhando, senge email i spam kana kwete, unoshandisa kupatsanura.
Kana iwe uchida kufanotaura shanduko inoenderera, semutengo wemba, unoshandisa regression. Clustering inoshandiswa kuisa pamwe chete inofananidzwa data zvinhu zvichienderana nezvavanofanana.
Kana tikatarisa muenzaniso wedu; dambudziko redu nderokuona kuti munhu anofarira zvemumhanzi musambo kubva pachirume nezera. Tichashandisa dhatabheti revanhu gumi nevasere pamuenzaniso uyu uye ruzivo rwezera ravo, murume kana mukadzi, uye musambo wavanoda.
2. Gadzirira data
Mushure mekunge watsanangura dambudziko, iwe unozofanirwa kugadzirira iyo data yekudzidzisa modhi. Izvi zvinosanganisira kuchenesa uye kugadzirisa data. Saka, kuti isu tive nechokwadi chekuti iri mufomati iyo iyo muchina kudzidza algorithm Unogona kushandisa.
Izvi zvinogona kusanganisira zviitiko zvakaita sekudzima hunhu husipo, kushandura dhata renhamba kudhata renhamba, uye kuyera kana kuenzanisa data kuti ive nechokwadi chekuti maitiro ese ari pachiyero chimwe.
Semuyenzaniso, aya ndiwo mabvisiro aunodzima hunhu husipo:
import pandas as pd
# Load the data into a pandas DataFrame
data = pd.read_csv('data.csv')
# Check for missing values
print(data.isnull().sum())
# Drop rows with missing values
data.dropna(inplace=True)
# Check that all missing values have been removed
print(data.isnull().sum())
Chinyorwa chidiki: Mumutsara o "import pandas as pd",
isu tinopinza raibhurari yePandas uye tinoigovera iyo alias "pd" kuita kuti zvive nyore kutarisa mabasa ayo uye zvinhu gare gare mukodhi.
Pandas imodhi inozivikanwa yePython yekushandura data uye kuongorora, kunyanya kana uchishanda neyakarongwa kana tabular data.
Mumuenzaniso wedu wekusarudza mhando dzemimhanzi. Tinotanga taunza dataset. Ndaritumidza kuti music.csv, zvisinei, unogona kuitumidza chero iwe waunoda.
Kugadzirira iyo data yekudzidzisa modhi yemuchina wekudzidza, tinoipatsanura kuita hunhu (zera uye murume kana mukadzi) uye zvinangwa (rudzi rwemimhanzi).
Isu tichawedzera kupatsanura iyo data kuita 80:20 kudzidziswa uye seti yekuyedza kuongorora mashandiro emodhi yedu uye kudzivirira kuwandisa.
# Import necessary libraries
import pandas as pd
from sklearn.model_selection import train_test_split
# Load data from CSV file/code>
music_data = pd.read_csv('music.csv')
# Split data into features and target
X = music_data.drop(columns=['genre'])
y = music_data['genre']
# Split data into training and testing sets
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)
3. Sarudza muenzaniso wekudzidza muchina.
Mushure mekunge wagadzirira iyo data, iwe unofanirwa kusarudza muchina-yekudzidza modhi inoenderana nebasa rako.
Kune akati wandei algorithms ekutora kubva, senge miti yesarudzo, logistic regression, kutsigira vector michina, neural network, uye mamwe. Iyo algorithm yaunosarudza ichatemwa nemhando yenyaya yauri kuyedza kupindura, rudzi rwe data raunaro, uye zvaunoda kuita.
Tichashandisa mucherechedzo wemuti wemucherechedzo wemuenzaniso uyu nekuti tiri kushanda nedambudziko rechikamu (kufanotaura dhata rechikamu).
# Import necessary libraries
from sklearn.tree import DecisionTreeClassifier
Heino tarisiro yemashandiro anoita Decision Tree Classifier:
4. Rovedza muenzaniso
Unogona kutanga kudzidzisa modhi kana wasarudza inogamuchirwa muchina-kudzidza algorithm. Izvi zvinosanganisira kushandisa iyo yakambogadzirwa data kudzidzisa iyo algorithm pamaitiro ekufanotaura pane nyowani, data risingaonekwe.
Iyo algorithm ichagadzirisa maparamendi ayo emukati panguva yekudzidziswa kuderedza mutsauko uripo pakati pezvainofanotaurwa kukosha uye chaiwo maitiro mune yekudzidziswa data. Huwandu hwedata hunoshandiswa pakudzidziswa, pamwe neiyo algorithm chaiyo paramita, zvese zvinogona kuve nemhedzisiro pakurongeka kweiyo mhedzisiro modhi.
Mumuenzaniso wedu chaiwo, zvino zvatasarudza nzira, tinogona kudzidzisa modhi yedu nedata rekudzidzisa.
# Train the decision tree classifier
model = DecisionTreeClassifier()
model.fit(X_train, y_train)
5. Ongorora muenzaniso
Mushure mekunge modhi yadzidziswa, inofanirwa kuongororwa pane nyowani data kuti ive nechokwadi chekuti ndeyechokwadi uye yakavimbika. Izvi zvinosanganisira kuyedza modhi nedata risina kushandiswa panguva yekudzidziswa uye kuenzanisa hunhu hwayo hunofungidzirwa nehunhu chaihwo mune data rebvunzo.
Ongororo iyi inogona kubatsira mukuona chero kukanganisa kwemodhi, sekukwirisa kana kusakwana, uye kunogona kutungamira kune chero kurongeka kungave kudikanwa.
Tichishandisa iyo data yekuyedza, isu tichaongorora kurongeka kwemuenzaniso wedu.
# Import necessary libraries
from sklearn.metrics import accuracy_score
# Predict the music genre for the test data
predictions = model.predict(X_test)
# Evaluate the model's accuracy
accuracy = accuracy_score(y_test, predictions)
print("Accuracy: ", accuracy)
Chibodzwa chechokwadi hachina kunyanya kushata parizvino. 🙂 Kuti uvandudze chibodzwa chako chechokwadi, unogona kugara uchichenesa iyo data zvakanyanya kana kuyedza akasiyana emuchina-wekudzidza modhi kuti uone kuti ndeipi inopa yakanyanya mamaki.
6. Gadzirisa muenzaniso
Kana iyo modhi inoshanda isina kukwana zvakakwana, unogona kuigadzirisa nekuchinja akasiyana algorithm paramita kana kuyedza nemaalgorithms matsva zvachose.
Iyi nzira inogona kusanganisira kuyedza mamwe mareti ekudzidza, kugadzirisa zvimiro zvenguva dzose, kana kushandura nhamba kana saizi yezvikamu zvakavanzwa muneural network.
7. Shandisa muenzaniso
Paunenge uchinge wafadzwa nekuita kweiyo modhi, unogona kutanga kuishandisa kugadzira fungidziro pane nyowani data.
Izvi zvinogona kusanganisira kudyisa data nyowani mumuenzaniso uye kushandisa iyo modhi yakadzidzwa paramita kugadzira fungidziro pane iyo data, kana kubatanidza iyo modhi kuita yakakura application kana sisitimu.
Tinogona kushandisa modhi yedu kugadzira fungidziro pane nyowani data mushure mekunge tafadzwa nekurongeka kwayo. Iwe unogona kuedza zvakasiyana-siyana zvehutano uye makore.
# Test the model with new data
new_data = [[25, 1], [30, 0]]
predictions = model.predict(new_data)
print("Predictions: ", predictions)
Putira
Isu tapedza kudzidzisa yedu yekutanga muchina kudzidza modhi.
Ndinovimba wazviwana zvichibatsira. Iwe unogona ikozvino kuedza kushandisa akasiyana muchina kudzidza modhi se Linear Regression kana Random Sango.
Kune akawanda dataset uye matambudziko mukati Kaggle kana iwe uchida kuvandudza yako coding uye kunzwisisa kwemuchina kudzidza.
Leave a Reply