-
Notifications
You must be signed in to change notification settings - Fork 9
Expand file tree
/
Copy pathMMMF.py
More file actions
39 lines (34 loc) · 1.66 KB
/
MMMF.py
File metadata and controls
39 lines (34 loc) · 1.66 KB
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
# -*- coding: utf-8 -*-
"""
Created on Thu Jul 19 10:06:21 2018
@author: Shinelon
"""
from VBR2016 import BPRMF
import numpy as np
class MMMF(BPRMF.BPRMF):
def __init__(self, corp, K, lambd, biasReg):
super(MMMF, self).__init__(corp, K, lambd, biasReg)
def updataFactors(self, user_id, pos_item_id, neg_item_id, learn_rate):
x_uij = self.beta_item[pos_item_id] - self.beta_item[neg_item_id]
x_uij += np.dot(self.gamma_user[user_id], self.gamma_item[pos_item_id]) - np.dot(self.gamma_user[user_id], self.gamma_item[neg_item_id])
deri = 1.0/(1+np.exp(x_uij))
if x_uij < 0:
deri = 1
else:
deri = 0
self.beta_item[pos_item_id] += learn_rate * (deri - self.biasReg * self.beta_item[pos_item_id])
self.beta_item[neg_item_id] += learn_rate * (-deri - self.biasReg * self.beta_item[neg_item_id])
for f in range(self.K):
w_uf = self.gamma_user[user_id][f]
h_if = self.gamma_item[pos_item_id][f]
h_jf = self.gamma_item[neg_item_id][f]
self.gamma_user[user_id][f] += learn_rate * ( deri * (h_if - h_jf) - self.lambd * w_uf)
self.gamma_item[pos_item_id][f] += learn_rate * ( deri * w_uf - self.lambd * h_if)
self.gamma_item[neg_item_id][f] += learn_rate * (-deri * w_uf - self.lambd / 10.0 * h_jf)
return
def tostring1(self):
print "MMMF__K_%d_lambda_%.2f_biasReg_%.2f"%(self.K, self.lambd, self.biasReg)
return
def tostring2(self):
print "<<< MMMF >>> Test AUC = %f, Test Std = %f\n"%(self.AUC_test, self.std)
return