trn_loss, trn_pred = fis.train(sess, trnData, trnLbls)
# Evaluate on validation set
val_pred, val_loss = fis.infer(sess, chkData, chkLbls)
if epoch % 10 == 0:
print("Train cost after epoch %i: %f" % (epoch, trn_loss))
if epoch == num_epochs - 1:
time_end = time.time()
print("Elapsed time: %f" % (time_end - time_start))
print("Validation loss: %f" % val_loss)
# Plot real vs. predicted
pred = np.vstack((np.expand_dims(trn_pred,
1), np.expand_dims(val_pred, 1)))
plt.figure(1)
plt.plot(mg_series)
plt.plot(pred)
trn_costs.append(trn_loss)
val_costs.append(val_loss)
# Plot the cost over epochs
plt.figure(2)
plt.subplot(2, 1, 1)
plt.plot(np.squeeze(trn_costs))
plt.title("Training loss, Learning rate =" + str(alpha))
plt.subplot(2, 1, 2)
plt.plot(np.squeeze(val_costs))
plt.title("Validation loss, Learning rate =" + str(alpha))
plt.ylabel('Cost')
plt.xlabel('Epochs')
# Plot resulting membership functions
fis.plotmfs(sess)
plt.show()
# Run an update step
trn_loss, trn_pred = fis.train(sess, trnData, trnLbls)
# Evaluate on validation set
val_pred, val_loss = fis.infer(sess, chkData, chkLbls)
if epoch % 10 == 0:
print("Train cost after epoch %i: %f" % (epoch, trn_loss))
if epoch == num_epochs - 1:
time_end = time.time()
print("Elapsed time: %f" % (time_end - time_start))
print("Validation loss: %f" % val_loss)
# Plot real vs. predicted
pred = np.vstack((np.expand_dims(trn_pred, 1), np.expand_dims(val_pred, 1)))
plt.figure(1)
plt.plot(mg_series)
plt.plot(pred)
trn_costs.append(trn_loss)
val_costs.append(val_loss)
# Plot the cost over epochs
plt.figure(2)
plt.subplot(2, 1, 1)
plt.plot(np.squeeze(trn_costs))
plt.title("Training loss, Learning rate =" + str(alpha))
plt.subplot(2, 1, 2)
plt.plot(np.squeeze(val_costs))
plt.title("Validation loss, Learning rate =" + str(alpha))
plt.ylabel('Cost')
plt.xlabel('Epochs')
# Plot resulting membership functions
fis.plotmfs(sess)
plt.show()
if trn_loss > 0.01:
continue
else:
pas = 1
# Plot the cost over epochs
plt.figure(2)
plt.subplot(2, 1, 1)
plt.plot(np.squeeze(trn_costs))
plt.title("Training loss, Learning rate =" + str(alpha))
plt.subplot(2, 1, 2)
plt.plot(np.squeeze(val_costs))
plt.title("Validation loss, Learning rate =" + str(alpha))
plt.ylabel('Cost')
plt.xlabel('Epochs')
# Plot resulting membership functions
rule_stats = fis.plotmfs(sess)
inp_list = [
"Seniority ", "Propensity ", "Size of Company ", "Contactibility "
]
action = "send brochure"
for r in rule_stats:
if abs(rule_stats[r]["center"]) > 0.05:
#print("rule %s" % r)
term = "\nrule %s: IF(" % r
for inp in range(4):
hb = rule_stats[r][inp]["high_bound"]
lb = rule_stats[r][inp]["low_bound"]
significance = ""
if lb >= 1 or hb <= 0: significance = "is not related "
elif hb == 1 and lb == 0: significance = "is in any value "