learning

Scikit-Learn 라이브러리

 

import pandas ad pd

df = pd.read_csv(C:/myCode/data/basketball_stat.csv')

data_df = df[0:100]

data_df.head()

 

from sklearn.model_selection import train_test_split

 

# 학습데이터 80%, 테스트데이터 20%로 분리

train, test = train_test_split(data_df, best_size=0.2)

train.shape

 

# 일부만 가져옴.

train_data_df = train[['3P',"BLK',"TRB']]

train_label_df = train[["Pos']]

tarin_data_df.head()

 

# KNN 라이브 추가

from sklearn.neighbors import KNeighborsClassfier

 

knn = KNeighborsClassifier(n_neighbors=3)

knn.fit(train_data, train_label)

 

pred_knn = knn.predict(test_data)

pred_knn

 

from sklearn import metrics

 

ac_score = metrics.accuracy_score(test_label, pred_knn)

print('accuracy:', ac_score)

 

# 모델 예측 정확도 확인

comparison = pd.DataFrame({'prediction':pred_knn, 'ground_truth':test_label})

comparison

 

# svm 모델 학습

from sklearn import svm

clf = svm.SVC(C = 1, gamma = 0.1) # gamma = 0.1은 가정

clif.fit(train_data, train_label)

 

# 테스트 데이터로 예측

pred_svm = clf.predict(test_data)

pred.svm

 

# 모델 예측 정확도 확인

from sklearn import metrics

 

ac_score = metrics.accuracy.score(test_label, pred_svm)

print('accuracy:', ac_score)

 

* Deep Learning

 

import tensorflow as tf

import pandas as pd

import numpy as np

 

df = pd.read_csv()

Data_set = df.values

Data_set

 

from sklearn.model_selection import train_test_split

 

# 데아터셋에서 학습데이터, 테스트데이터를 분리

train, test = train_test_split(Data_set, test_size = 0.2)

 

# 속성과 클래스 분리

X_train = train[:, 0:17]

Y_train = train[:, 17]

X_train

 

from keras.models import Sequential

from keras.layers import Dense

 

model = Sequential()

 

model.add(Dense(30, input dim=17, activation='relu'))

model.add(dense(1, activation='sigmoid'))

 

model.compile(loss='mean_squared_error', optimizer = 'adam', metrics=['accuracy'])