목록LSTM (9)

잠토의 잠망경

[ML] LSTM - Univariate Bidirectional LSTM Models

GITHUB https://github.com/yiwonjae/Project_Lotto/blob/master/Book_001/p130.py 0. 목표 ''' data [10, 20, 30, 40, 50, 60, 70, 80, 90] X(input), y(output) 10, 20, 30, 40 20, 30, 40, 50 30, 40, 50, 60 ''' 1. DATA raw_seq = np.asarray([10, 20, 30, 40, 50, 60, 70, 80, 90]) 2. DATA 정제 X import numpy as np from numpy import ndarray def split_sequence(sequence:ndarray, n_steps:int)->(ndarray, ndarray): x =..
공부/Python 2020. 2. 16. 17:42
[ML] LSTM - Univariate stack LSTM Models

GITHUB https://github.com/yiwonjae/Project_Lotto/blob/master/Book_001/p128.py 0. 목표 ''' data [10, 20, 30, 40, 50, 60, 70, 80, 90] X(input), y(output) 10, 20, 30, 40 20, 30, 40, 50 30, 40, 50, 60 ''' 1. DATA raw_seq = np.asarray([10, 20, 30, 40, 50, 60, 70, 80, 90]) 2. DATA 정제 X import numpy as np from numpy import ndarray def split_sequence(sequence:ndarray, n_steps:int)->(ndarray, ndarray): x =..
공부/Python 2020. 2. 16. 17:37
[ML] LSTM - Univariate LSTM Models

GITHUB https://github.com/yiwonjae/Project_Lotto/blob/master/Book_001/p127_LSTM.py 0. 목표 ''' data [10, 20, 30, 40, 50, 60, 70, 80, 90] X(input), y(output) 10, 20, 30, 40 20, 30, 40, 50 30, 40, 50, 60 ''' 1. DATA raw_seq = np.asarray([10, 20, 30, 40, 50, 60, 70, 80, 90]) 2. DATA 정제 X import numpy as np from numpy import ndarray def split_sequence(sequence:ndarray, n_steps:int)->(ndarray, ndarray)..
공부/Python 2020. 2. 16. 17:32
[ML] Multiple Input Multi-step Output

GITHUB https://github.com/yiwonjae/Project_Lotto/blob/master/Book_001/p113.py 0. 목표 input 多, output 多step 1. DATA in_seq1 = np.asarray([10, 20, 30, 40, 50, 60, 70, 80, 90]) in_seq2 = np.asarray([15, 25, 35, 45, 55, 65, 75, 85, 95]) out_seq = np.asarray([in_seq1[i]+in_seq2[i] for i in range(len(in_seq1))]) in_seq1:ndarray = in_seq1.reshape((len(in_seq1), 1)) in_seq2:ndarray = in_seq2.reshape((len..
공부/Python 2020. 2. 16. 10:12
[ML] Multi-output Parallel Series

GITHUB https://github.com/yiwonjae/Project_Lotto/blob/master/Book_001/p107.py 0. 목표 feature의 수에 맞추서 output을 산출하는 방법이다. 1. DATA in_seq1 = np.array([10, 20, 30, 40, 50, 60, 70, 80, 90]) in_seq2 = np.array([15, 25, 35, 45, 55, 65, 75, 85, 95]) out_seq = np.array([in_seq1[i]+in_seq2[i] for i in range(len(in_seq1))]) in_seq1 = in_seq1.reshape((len(in_seq1), 1)) in_seq2 = in_seq2.reshape((len(in_seq2)..
공부/Python 2020. 2. 15. 21:29
[ML] Multiple Parallel Series

GITHUB https://github.com/yiwonjae/Project_Lotto/blob/master/Book_001/p104.py 0. 목표 이전 값들을 갖고 한단계 앞 값을 예측한다. 1. DATA in_seq1 = np.array([10, 20, 30, 40, 50, 60, 70, 80, 90]) in_seq2 = np.array([15, 25, 35, 45, 55, 65, 75, 85, 95]) out_seq = np.array([in_seq1[i]+in_seq2[i] for i in range(len(in_seq1))]) in_seq1 = in_seq1.reshape((len(in_seq1), 1)) in_seq2 = in_seq2.reshape((len(in_seq2), 1)) out_..
공부/Python 2020. 2. 15. 17:11
[ML] Multiple Input Series - Multivariate CNN Models

GITHUB https://github.com/yiwonjae/Project_Lotto/blob/master/Book_001/p096.py 0. 목표 data가 두가지 종류가 있다. 현재 포함 과거 data 총 3단계를 바탕으로 예측되는 data를 산출한다. input data 10, 15 20, 25 30, 35 prediction data 65 1. DATA in_seq1 = np.asarray([10, 20, 30, 40, 50, 60, 70, 80, 90]) in_seq2 = np.asarray([15, 25, 35, 45, 55, 65, 75, 85, 95]) out_seq = np.asarray([in_seq1[i]+in_seq2[i] for i in range(len(in_seq1))]) i..
공부/Python 2020. 2. 15. 11:20
[ML] Univariate CNN Models

GITHUB https://github.com/yiwonjae/Project_Lotto/blob/master/Book_001/p091.py 0. 목표 연속된 하나의 sequcence에서 다음 하나의(univariate) 값을 예측한다. data : [10, 20, 30, 40, 50, 60, 70, 80, 90] x y 10 20 30 40 20 30 40 50 30 40 50 60 1. DATA raw_seq = [10, 20, 30, 40, 50, 60, 70, 80, 90] n_steps = 3 2. DATA 정제 X def splite_sequence(sequence:list, n_steps:int)->(ndarray, ndarray): x, y = [], [] for i in range(len(..
공부/Python 2020. 2. 15. 10:43
Prev 1 2 Next