forked from Evo-ML/EvoloPy-NN
-
Notifications
You must be signed in to change notification settings - Fork 0
/
MVO.py
177 lines (108 loc) · 4.33 KB
/
MVO.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
# -*- coding: utf-8 -*-
"""
Created on Wed May 11 17:06:34 2016
@author: hossam
"""
import random
import numpy
import time
import math
import sklearn
from numpy import asarray
from sklearn.preprocessing import normalize
from solution import solution
def normr(Mat):
"""normalize the columns of the matrix
B= normr(A) normalizes the row
the dtype of A is float"""
Mat=Mat.reshape(1, -1)
# Enforce dtype float
if Mat.dtype!='float':
Mat = asarray(Mat,dtype=float)
# if statement to enforce dtype float
B = normalize(Mat,norm='l2',axis=1)
B=numpy.reshape(B,-1)
return B
def randk(t):
if (t%2)==0:
s=0.25
else:
s=0.75
return s
def RouletteWheelSelection(weights):
accumulation = numpy.cumsum(weights)
p = random.random() * accumulation[-1]
chosen_index = -1;
for index in range (0, len(accumulation)):
if (accumulation[index] > p):
chosen_index = index;
break;
choice = chosen_index;
return choice
def MVO(objf,lb,ub,dim,N,Max_time,trainInput,trainOutput,net):
#def MVO(objf,lb,ub,dim,N,Max_time):
"parameters"
#dim=30
#lb=-100
#ub=100
WEP_Max=1;
WEP_Min=0.2
#Max_time=1000
#N=50
Universes=numpy.random.uniform(0,1,(N,dim)) *(ub-lb)+lb
Sorted_universes=numpy.copy(Universes)
convergence=numpy.zeros(Max_time)
Best_universe=[0]*dim;
Best_universe_Inflation_rate= float("inf")
s=solution()
Time=1;
############################################
print("MVO is optimizing \""+objf.__name__+"\"")
timerStart=time.time()
s.startTime=time.strftime("%Y-%m-%d-%H-%M-%S")
while (Time<Max_time+1):
"Eq. (3.3) in the paper"
WEP=WEP_Min+Time*((WEP_Max-WEP_Min)/Max_time)
TDR=1-(math.pow(Time,1/6)/math.pow(Max_time,1/6))
Inflation_rates=[0]*len(Universes)
for i in range(0,N):
Universes[i,:]=numpy.clip(Universes[i,:], lb, ub)
Inflation_rates[i]=objf(Universes[i,:],trainInput,trainOutput,net);
if Inflation_rates[i]<Best_universe_Inflation_rate :
Best_universe_Inflation_rate=Inflation_rates[i]
Best_universe=numpy.array(Universes[i,:])
sorted_Inflation_rates = numpy.sort(Inflation_rates)
sorted_indexes = numpy.argsort(Inflation_rates)
for newindex in range(0,N):
Sorted_universes[newindex,:]=numpy.array(Universes[sorted_indexes[newindex],:])
normalized_sorted_Inflation_rates=numpy.copy(normr(sorted_Inflation_rates))
Universes[0,:]= numpy.array(Sorted_universes[0,:])
for i in range(1,N):
Back_hole_index=i
for j in range(0,dim):
r1=random.random()
if r1<normalized_sorted_Inflation_rates[i]:
White_hole_index=RouletteWheelSelection(-sorted_Inflation_rates);
if White_hole_index==-1:
White_hole_index=0;
White_hole_index=0;
Universes[Back_hole_index,j]=Sorted_universes[White_hole_index,j];
r2=random.random()
if r2<WEP:
r3=random.random()
if r3<0.5:
Universes[i,j]=Best_universe[j]+TDR*((ub-lb)*random.random()+lb) #random.uniform(0,1)+lb);
if r3>0.5:
Universes[i,j]=Best_universe[j]-TDR*((ub-lb)*random.random()+lb) #random.uniform(0,1)+lb);
convergence[Time-1]=Best_universe_Inflation_rate
if (Time%1==0):
print(['At iteration '+ str(Time)+ ' the best fitness is '+ str(Best_universe_Inflation_rate)]);
Time=Time+1
timerEnd=time.time()
s.endTime=time.strftime("%Y-%m-%d-%H-%M-%S")
s.executionTime=timerEnd-timerStart
s.convergence=convergence
s.optimizer="MVO"
s.objfname=objf.__name__
s.bestIndividual=Best_universe
return s