기계학습기초_4주차과제.py

# -*- coding: utf-8 -*-
"""백민하201704045.ipynb

Automatically generated by Colaboratory.

Original file is located at
    https://colab.research.google.com/drive/1RUdNBwZ1A8GLPdlDI5CQmupQdPqlfJ7i
"""

import matplotlib.pyplot as plt
import numpy as np
from sklearn.neighbors import KNeighborsClassifier
from sklearn.model_selection import train_test_split

fst_grp = [1, 2, 3, 4, 5]
snd_grp = [6, 7, 8, 9, 10]

num_grp = fst_grp + snd_grp

num_data = np.column_stack(([num_grp]))
num_target = np.concatenate((np.ones(5), np.zeros(5)))

kn3s = []
kn5s = []
for i in range(0, 1000):
  trn_ipt, tst_ipt, trn_tgt,  tst_tgt = train_test_split(num_data, num_target, test_size = 0.4, stratify=num_target)
  kn3 = KNeighborsClassifier(3)
  kn3.fit(trn_ipt, trn_tgt)
  kn3s.append(kn3.score(tst_ipt, tst_tgt))

  kn5 = KNeighborsClassifier(5)
  kn5.fit(trn_ipt, trn_tgt)
  kn5s.append(kn5.score(tst_ipt, tst_tgt))

print(f'k=3일때 평균: {np.mean(kn3s)} / 표준편차: {np.std(kn3s)}\nk=5일때 평균: {np.mean(kn5s)} / 표준편차: {np.std(kn5s)}')