?????? ????, ???? & Tensorflow
Download as pptx, pdf6 likes2,710 views
- ????? ??? ?????? - ? ?? ????? ???? - ????? Tensorflow ?? ????
![? ??? ?? ???? [????]
? ???
- ???
- ?? ?? ?? -> ???/??? ???
? ???
- ??, ???, ????
- ? ?? ?? ?? -> ? ???/??? ???
- ????
? ??](https://image.slidesharecdn.com/random-170425072127/85/Tensorflow-8-320.jpg)
![?????? ?? ???? Cycle
?????
OP? ??
????
OP? ??/???
??
??/??
??? ??
? ???? - [??2 ? ?????6]
110? ?? ? 105? ?? (95.45%)
? ???? C [??? ??6]
131? ??? ? 57? ??(43.51%)
- ??? ??4 : 120? ? 61? ?? (50.83%)
- ??? ??5 : 126? ? 74? ?? (58.73%)](https://image.slidesharecdn.com/random-170425072127/85/Tensorflow-13-320.jpg)
![Tensorflow ??
# Lab 4 Multi-variable linear regression
import tensorflow as tf
import numpy as np
tf.set_random_seed(378) # for reproducibility
xy = np.loadtxt('databalancing.csv', delimiter=',', dtype=np.float32)
#xy = np.loadtxt('databalancing_YTV.csv', delimiter=',', dtype=np.float32)
x_data = xy[:, 0:-1]
#x_data = (xy[:, 0:-1] - xy[:, 0:-1].mean()) / xy[:, 0:-1].std()
y_data = xy[:, [-1]]
# Make sure the shape and data are OK
#print(x_data.shape, x_data, len(x_data))
#print(y_data.shape, y_data)
# placeholders for a tensor that will be always fed.
X = tf.placeholder(tf.float32, shape=[None, 9])
Y = tf.placeholder(tf.float32, shape=[None, 1])
#????
W = tf.Variable(tf.random_normal([9, 1]), name='weight')
b = tf.Variable(tf.random_normal([1]), name='bias')
#????
#W = tf.Variable(tf.random_uniform([9, 1], minval=0.3, maxval=0.5), name='weight')
#b = tf.Variable(tf.random_normal([1]), name='bias')
# Hypothesis using sigmoid: tf.div(1., 1. + tf.exp(tf.matmul(X, W)))
hypothesis = tf.sigmoid(tf.matmul(X, W) + b)
# cost/loss function
cost = -tf.reduce_mean(Y * tf.log(tf.clip_by_value(hypothesis, 1e-8, 1.)) +
(1 - Y) * tf.log(tf.clip_by_value(1 - hypothesis, 1e-8, 1.)))
#cost = -tf.reduce_mean(Y * tf.log(hypothesis) +
# (1 - Y) * tf.log(1 - hypothesis))
# Accuracy computation
# True if hypothesis>0.5 else False](https://image.slidesharecdn.com/random-170425072127/85/Tensorflow-20-320.jpg)
?????? ????, ???? & Tensorflow
- 1. ???? ????? ???? ???? ????? Tensorflow ??? : ??? ( thx@daum.net )
- 2. ?? ? ???? = ??? ????. ????
- 3. ????, ????, ??? ?? ? ??? ? ???? ? ????
- 4. ??.. ??? AI? ????
- 5. ?????? ?? ? ??? ???.
- 7. ?? ? ???? ? ?????, ?? IP ??，???? ??? ???? - (?????) ? ???，??，??, AI?? ?? ??，??，??? ?? (????) ? ????&?? ??? ? ?? ???? ???? ??? ?? (????) ? ?????? ???? ?? C ??, ??, ??? AI ?? ?? ?? ?? (?? ???) ? ??? AI，????´ '??? ???' ??? C (Chosunbiz)
- 8. ? ??? ?? ???? [????] ? ??? - ??? - ?? ?? ?? -> ???/??? ??? ? ??? - ??, ???, ???? - ? ?? ?? ?? -> ? ???/??? ??? - ???? ? ??
- 9. ???? ??? ? ????? ??? ?? ??? ??????? ????? ?? ?, ???? ?????? ???? ??? ???? ? ?? = ??? ??
- 10. ?? ???? ?????? ? ???? ?? : Korea Games Conference ????? ??????
- 11. ?? ???? ?????? ? ?? ????! ??? = DAU * PU% * ARPPU ? DAU = SUM ( NRU * Retention ) ? NRU? ?? = ? ? Retention = ??? ?? ? ???? ? ??! ＾???￣ ?? ????!! (???) ??? ?? ?? ?? : ???!! ???? : ?????!!
- 12. ?? 』 ??? But, ?? ? ??? ÷ ???? ?????????!!
- 13. ?????? ?? ???? Cycle ????? OP? ?? ???? OP? ??/??? ?? ??/?? ??? ?? ? ???? - [??2 ? ?????6] 110? ?? ? 105? ?? (95.45%) ? ???? C [??? ??6] 131? ??? ? 57? ??(43.51%) - ??? ??4 : 120? ? 61? ?? (50.83%) - ??? ??5 : 126? ? 74? ?? (58.73%)
- 14. ML ?? ????. ??? ?? ??? (feat. ? ?? ?? / chorus. MS The ??)
- 16. Tensorflow ?? ? ???? ?? ?? ???? ????? ? Python / C++ ?? ? CPU / GPU ???? ? ??? ???? ?? ???? ??
- 17. Tensorflow ???? ? Python 3.5.3 / Anaconda 4.2.0 ? IDE : Pycharm ? ????? : tensorflow, plot
- 18. ???? ? ?? ? ???? 1. ?? ? ??? ??, ??? ??/??(??????) 2. ?? ???? ? ???? ??? ??? ?? ? ?? 1. ??? ?? ? ??? ?? ＾??? ??? ??? ??/?? ??＾
- 19. ????? ?? ? ??? ??/?? ?? ? ??? ?? * ??? = ?? ?????? ?? ???? < 2.5 ? ?? ??? ? ?? ? ??? ?? ?? ??, ??, ??? (??? ? ?? ??? ?? C unknown) ? ????? : tensorflow, plot
- 20. Tensorflow ?? # Lab 4 Multi-variable linear regression import tensorflow as tf import numpy as np tf.set_random_seed(378) # for reproducibility xy = np.loadtxt('databalancing.csv', delimiter=',', dtype=np.float32) #xy = np.loadtxt('databalancing_YTV.csv', delimiter=',', dtype=np.float32) x_data = xy[:, 0:-1] #x_data = (xy[:, 0:-1] - xy[:, 0:-1].mean()) / xy[:, 0:-1].std() y_data = xy[:, [-1]] # Make sure the shape and data are OK #print(x_data.shape, x_data, len(x_data)) #print(y_data.shape, y_data) # placeholders for a tensor that will be always fed. X = tf.placeholder(tf.float32, shape=[None, 9]) Y = tf.placeholder(tf.float32, shape=[None, 1]) #???? W = tf.Variable(tf.random_normal([9, 1]), name='weight') b = tf.Variable(tf.random_normal([1]), name='bias') #???? #W = tf.Variable(tf.random_uniform([9, 1], minval=0.3, maxval=0.5), name='weight') #b = tf.Variable(tf.random_normal([1]), name='bias') # Hypothesis using sigmoid: tf.div(1., 1. + tf.exp(tf.matmul(X, W))) hypothesis = tf.sigmoid(tf.matmul(X, W) + b) # cost/loss function cost = -tf.reduce_mean(Y * tf.log(tf.clip_by_value(hypothesis, 1e-8, 1.)) + (1 - Y) * tf.log(tf.clip_by_value(1 - hypothesis, 1e-8, 1.))) #cost = -tf.reduce_mean(Y * tf.log(hypothesis) + # (1 - Y) * tf.log(1 - hypothesis)) # Accuracy computation # True if hypothesis>0.5 else False
- 21. ?? ? ???? ?? ?? : 90.2 %
- 22. ?? ? ???? ? ?? ???? ???? ?? ??? ? ?? ???? ?? ? ?? ??? ??? ?? ?? ???? ??? : 90% ? 75% ? CNN, RNN, Reinforce Learning ? ???? ?? ? learning rate, ?? ? ? ???
- 23. Q & A
- 24. ?????