경기도미래기술학교 AI개발자 부트캠프 31일차 TIL- Continuous Descriptive Features

https://process-mining.tistory.com/42

Decision Tree(의사 결정 나무)란? (Decision tree 설명)

 

import numpy as np
# elevation_list = [300,1200,1500,3000,3900,4450,5000]

# vegetation_list = [rip,cha,rip,cha,cha,con,con]

#4개로 나뉜다.
# rip/cha/rip/cha,cha/con,con

# 750

n_number = 2

target_level_differ_1 = (300+1200)//n_number
target_level_differ_2 = (1200+1500)//n_number
target_level_differ_3 = (1500+3000)//n_number
target_level_differ_4 = (3900+4450)//n_number

print('target_level_differ_1 : ',target_level_differ_1)
print('target_level_differ_2 : ',target_level_differ_2)
print('target_level_differ_3 : ',target_level_differ_3)
print('target_level_differ_4 : ',target_level_differ_4)

print('-------------##root node##-------------')
root_node = (-3/7*np.log2(3/7)) + ((-2/7*np.log2(2/7))*2)
print('root_entropy: ',root_node)
print('-------------750 start-------------')
low_node_750 = (-1/1*np.log2(1/1))

print('low_node_750: ', low_node_750)

low_weight_node_750 = 1/7*low_node_750

print('low_weight_node_750: ',low_weight_node_750)

high_node_750 = (-3/6*np.log2(3/6))+(-2/6*np.log2(2/6))+(-1/6*np.log2(1/6))

print('high_node_750: ', high_node_750)

high_weight_node_750 = 6/7*high_node_750

print('high_weight_node_750: ',high_weight_node_750)

ig_750 = root_node - (high_weight_node_750+low_weight_node_750)

print('lower than 750 ig: ',ig_750)

print('-------------1350 start-------------')
low_node_1350 = (-1/2*np.log2(1/2))*2

print('low_weight_node_1350: ',low_node_1350)

low_weight_node_1350 = 2/7*low_node_1350

print('low_weight_node_1350: ',low_weight_node_1350)

high_node_1350 = (-2/5*np.log2(2/5))+(-2/5*np.log2(2/5))+(-1/5*np.log2(1/5))

print('high_node_1350: ', high_node_1350)

high_weight_node_1350 = 5/7*high_node_1350

print('high_weight_node_1350: ',high_weight_node_1350)

ig_1350 = root_node - (low_weight_node_1350+high_weight_node_1350)

print('lower than 1350 ig: ',ig_1350)

print('-------------2250 start-------------')

low_node_2250 = ((-2/3*np.log2(2/3))+(-1/3*np.log2(1/3)))

print('low_node_2250: ', low_node_2250)

low_weight_node_2250 = 3/7*low_node_2250

print('low_weight_node_2250: ',low_weight_node_2250)

high_node_2250 = ((-1/2*np.log2(1/2))*2)

print('high_node_2250: ', high_node_2250)

high_weight_node_2250 = 4/7*high_node_2250

print('high_weight_node_2250: ',high_weight_node_2250)



ig_2250 = root_node - (low_weight_node_2250+high_weight_node_2250)

print('lower than 2250 ig: ',ig_2250)

print('-------------4150 start-------------')

low_node_4150 = (-2/5*np.log2(2/5))+(-3/5*np.log2(3/5))

print('low_node_4150: ', low_node_4150)

low_weight_node_4150 = 5/7*low_node_4150

print('low_weight_node_4150: ',low_weight_node_4150)

high_node_4150 = ((-1*np.log2(1))*2)

print('high_node_4150: ', high_node_4150)

high_weight_node_4150 = 2/7*high_node_4150

print('high_weight_node_4150: ',high_weight_node_4150)

ig_4150 = root_node - (low_weight_node_4150+high_weight_node_4150)

print('lower than 4150 ig: ',ig_4150)

print('-------------4150 seperated -------------')
elevation_4150_node = (-3/5*np.log2(3/5)) + ((-2/5*np.log2(2/5)))
print('elevation_4150_node: ',elevation_4150_node)
print('-------------stream seperated test -------------')
stream_true_node_weight = 3/5 * ((-2/3*np.log2(2/3)) + ((-1/3*np.log2(1/3))))

print('stream_true_node_weight: ',stream_true_node_weight)

stream_false_node_weight = 2/5*(-1*np.log2(1))

print('stream_false_node_weight: ',stream_false_node_weight)

ig = elevation_4150_node-(stream_true_node_weight+stream_false_node_weight)
print('stream ig: ',ig)

print('-------------slope seperated test -------------')

slope_moderate_node_weight = 1/5*(-1*np.log2(1))

print('slope_moderate_node_weight: ',slope_moderate_node_weight)

slope_steep_node_weight = 4/5*((-3/4*np.log2(3/4))+(-1/4*np.log2(1/4)))

print('slope_steep_node_weight: ',slope_steep_node_weight)

ig = elevation_4150_node-(slope_moderate_node_weight+slope_steep_node_weight)
print('slope ig: ',ig)

print('-------------elevation seperated test -------------')

print('target_level_differ_1 : ',target_level_differ_1)
print('target_level_differ_2 : ',target_level_differ_2)
print('target_level_differ_3 : ',target_level_differ_3)

print('-------------elevation seperated 750 test -------------')
low_weight_node_750 = 1/5*low_node_750

print('low_weight_node_750: ',low_weight_node_750)

high_node_750 = (-3/4*np.log2(3/4))+(-1/4*np.log2(1/4))

print('high_node_750: ', high_node_750)

high_weight_node_750 = 4/5*high_node_750

print('high_weight_node_750: ',high_weight_node_750)

ig_750 = elevation_4150_node - (high_weight_node_750+low_weight_node_750)

print('lower than 750 ig: ',ig_750)

print('-------------elevation seperated 1350 test -------------')

low_weight_node_1350 = 2/5*low_node_1350

print('low_weight_node_1350: ',low_weight_node_1350)

high_node_1350 = (-1/3*np.log2(1/3))+(-2/3*np.log2(2/3))

print('high_node_1350: ', high_node_1350)

high_weight_node_1350 = 3/5*high_node_1350

print('high_weight_node_1350: ',high_weight_node_1350)

ig_1350 = elevation_4150_node - (low_weight_node_1350+high_weight_node_1350)

print('lower than 1350 ig: ',ig_1350)

print('-------------elevation seperated 2250 test -------------')

low_weight_node_2250 = 3/5*low_node_2250

print('low_weight_node_2250: ',low_weight_node_2250)

high_node_2250 = ((-1/1*np.log2(1/1))*2)

print('high_node_2250: ', high_node_2250)

high_weight_node_2250 = 2/5*high_node_2250

print('high_weight_node_2250: ',high_weight_node_2250)

ig_2250 = elevation_4150_node - (low_weight_node_2250+high_weight_node_2250)

print('lower than 2250 ig: ',ig_2250)


print('-------------3rd v1 : elevation seperated 2250 -------------')
elevation_2nd_2250_node = (-2/3*np.log2(2/3)) + (-1/3*np.log2(1/3))
print('elevation_2nd_2250_node: ',elevation_2nd_2250_node)

print('-------------3rd : stream test --------------')

stream_3rd_false_node_weight = 0

print('stream_3rd_false_node_weight: ',stream_false_node_weight)

ig = elevation_2nd_2250_node-(stream_3rd_false_node_weight)
print('stream ig: ',ig)

print('-------------3rd : slope test --------------')

slope_3rd_moderate_node_weight = 0

print('slope_3rd_moderate_node_weight: ',slope_3rd_moderate_node_weight)

slope_3rd_steep_node_weight = 2/3*((-1/2*np.log2(1/2))+(-1/2*np.log2(1/2)))

print('slope_steep_node_weight: ',slope_steep_node_weight)

ig = elevation_2nd_2250_node-(slope_3rd_moderate_node_weight+slope_3rd_steep_node_weight)
print('slope ig: ',ig)

print('-------------3rd : elevation test start--------------')
print('target_level_differ_1 : ',target_level_differ_1)
print('target_level_differ_2 : ',target_level_differ_2)

print('-------------3rd : 750 test --------------')

print('750 ig: ',ig)

print('-------------3rd : 1350 test --------------')

print('1350 ig: ',ig)

print('-------------3rd v2 : stream -------------')

true_3rd_node = (-2/3*np.log2(2/3)) + (-1/3*np.log2(1/3))
print('true_3rd_node: ',true_3rd_node)

print('-------------3rd v2 : slope test -------------')

 

import numpy as np

#season 때 많이 타냐, day 때 많이 타는지 분석하고 회귀분석

def calculate_variance(data):
n = len(data)
mean = sum(data) / n
squared_deviations = [(x - mean) ** 2 for x in data]
variance = sum(squared_deviations) / n
return variance


total_data = [800,826,900,2100,4740,4900,3000,5800,6200,2910,2880,2820]

n = len(total_data)
mean = sum(total_data)/n

#편차를 제곱한다.
squared_deviations = [(x-mean)**2 for x in total_data]
#평균한다.
#분산은 편차 제곱의 평균.
total_variance = sum(squared_deviations)/n

print(total_variance)

winter_data = [800, 826, 900]
spring_data = [2100, 4740, 4900]
summer_data = [3000, 5800, 6200]
autumn_data = [2910, 2880, 2820]

winter_variance = calculate_variance(winter_data)
print('winter_variance:', winter_variance)

spring_variance = calculate_variance(spring_data)
print('spring_variance:', spring_variance)

summer_variance = calculate_variance(summer_data)
print('summer_variance:', summer_variance)

autumn_variance = calculate_variance(autumn_data)
print('autumn_variance:', autumn_variance)

season_variance = (1/4 * spring_variance) + (1/4 * summer_variance) + (1/4 * autumn_variance) + (1/4 * winter_variance)
print('season_variance:', season_variance)

print('------------####day start#####-------------')

false_data = [800,826,2100,3000,2910,2880]
true_data = [900,4740,4900,5800,6200,2820]

false_variance = calculate_variance(false_data)
print('false_variance:', false_variance)

true_variance = calculate_variance(true_data)
print('true_variance:', true_variance)

day_variance = (1/2 * false_variance) + (1/2 * true_variance)
print('day_variance:', day_variance)