sample notebooks

[ MAPE 구하는 함수 ]

https://gist.github.com/amanahuja/6315882

[ 아나콘다에서 윈도우즈 텐서플로 설치 ]

https://tensorflow.blog/%EC%9C%88%EB%8F%84%EC%9A%B0%EC%A6%88%EC%97%90-%EC%95%84%EB%82%98%EC%BD%98%EB%8B%A4-%ED%85%90%EC%84%9C%ED%94%8C%EB%A1%9C%EC%9A%B0-%EC%84%A4%EC%B9%98%ED%95%98%EA%B8%B0/

[ multivariate time series forecasting using lstm :: references ]

Multivariate Time Series Forecasting with LSTMs in Keras

https://machinelearningmastery.com/multivariate-time-series-forecasting-lstms-keras/?fbclid=IwAR0DG94DGhOUpCVmFlQFAVX_sTgunfqpZscieeRKH0dz41rcmSdmDgM0Qn8

Multivariate Time Series Forecasting with LSTMs in Keras

https://hk.saowen.com/a/620ee85e064d178fecd830840c37f5d98aca7094723de125d49d5ed085b57243?fbclid=IwAR0LNsnBie8zBOr9ic-WrqflOIyhh4ED0XWbnVg-F7J40CONrk8MyxZYNx8

Time-Series Modeling with Neural Networks at Uber

https://forecasters.org/wp-content/uploads/gravity_forms/7-c6dd08fee7f0065037affb5b74fec20a/2017/07/Laptev_Nikolay_ISF2017.pdf?fbclid=IwAR0fNC7nPtPa9q9Fw6QURi2SW2GBXAqRbScFvVNo0AQ5o_gNW3YeVR40xPM

LSTMs for Human Activity Recognition

https://github.com/guillaume-chevalier/LSTM-Human-Activity-Recognition?fbclid=IwAR0LNsnBie8zBOr9ic-WrqflOIyhh4ED0XWbnVg-F7J40CONrk8MyxZYNx8 

Correlated Time Series Forecasting using Deep Neural Networks: A Summary of Results 

(AE vs Convolutional RNN) 

https://arxiv.org/pdf/1808.09794.pdf

[ Forecasting Multivariate Time Series Data Using Neural Networks ]

Forecasting Multivariate Time Series Data Using... (PDF) : 구글에서 위 제목으로 검색

:: XGBoost, CatBoost

전용준_리비젼_xgbcatb_201810.pdf

# SVM feature importance plotting

# https://medium.com/@aneesha/visualising-top-features-in-linear-svm-with-scikit-learn-and-matplotlib-3454ab18a14d

#========================

import numpy as np

import matplotlib.pyplot as plt

import pandas as pd

# Read data

d0 = pd.read_csv("C:/Users/KDATA/Desktop/Y00/cat_.csv", encoding = "euc_kr")

# d0 = pd.read_csv("C:\Users\KDATA\Desktop\Y00\cat_.csv")

print(d0.head())

# d19 = d1['Q53A4_2'][:20]

# print(d19)

# replace string value of null to np.Nan

d1['Q53A4_2'].replace('#NULL!', np.NaN, inplace=True) 

d19 = d1['Q53A4_2'][d1['Q53A4_2'].notnull()]

print(d19[:10])

# d19 = d19[~np.isnan(d19)]

print(d19[:10])

d191 = list(map(int, np.array(d19)))

plt.hist(d19, bins=20)

plt.show()

autoencoder_01.ipynb

DavisEDA.ipynb

EDA_scatterplot_practice.ipynb

embedding_01.ipynb

movies_vec.ipynb

RF_bank.ipynb

tfldl01.ipynb

tfldl01_cont_tgt.ipynb

tfldl01_cont_tgt_ML18.ipynb

[ AutoEncoder Sample and Beijing Data Set ]

AE_03.ipynb

Beijing2017_HourlyPM25.csv

[ DT 의미 파악 예제 - Boston data ]

Boston 디시전트리 예제.ipynb

[ correlation heatmap 예제 ]

https://www.linkedin.com/pulse/generating-correlation-heatmaps-seaborn-python-andrew-holt  

[ 출력된 챠트 해상도 높이기 : 참고 ] 

https://stackoverflow.com/questions/12192661/matplotlib-increase-resolution-to-see-details 

import pylab as pl
pl.figure(figsize=(7, 7))  # Don't create a humongous figure
pl.annotate(..., fontsize=1, ...)   # probably need the annotate line *before* savefig
pl.savefig('test.pdf', format='pdf')   # no need for DPI setting, assuming the fonts and figures are all vector based

[ 스캐터플롯에 대각선 추가용 참고]

# regression part

from scipy import stats

slope, intercept, r_value, p_value, std_err = stats.linregress(x,y1)

line = slope*x+intercept

plt.plot(x, line, '--')

막대그래프 그리기 

plt.bar(np.arange(3),[1,2,3])

plt.xticks(np.arange(3), names)

[item-CF example]

Item_CF.ipynb

[ ARIMA 예제 ]

https://machinelearningmastery.com/arima-for-time-series-forecasting-with-python/  

https://medium.com/@josemarcialportilla/using-python-and-auto-arima-to-forecast-seasonal-time-series-90877adff03c

전용준_리비젼_xgbcatb_201810.pdf

0.44MB

movies_vec.ipynb

0.67MB

EDA_scatterplot_practice.ipynb

0.15MB

RF_bank.ipynb

1.16MB

embedding_01.ipynb

0.02MB

tfldl01_cont_tgt_ML18.ipynb

0.63MB

AE_03.ipynb

0.08MB

DavisEDA.ipynb

0.08MB

tfldl01_cont_tgt.ipynb

0.86MB

autoencoder_01.ipynb

0.17MB

Beijing2017_HourlyPM25.csv

0.19MB

Item_CF.ipynb

0.11MB

Boston 디시전트리 예제.ipynb

0.1MB

AE_03.ipynb

0.04MB

tfldl01.ipynb

1.89MB