cas-pml/SL/aufgaben/template/2_Code/1.1 Feature Engineering - Einfuehrung.ipynb

nicht als Notebook verteilen

Feature Engineering

Feature Engineering - Einführung

Abgrenzungen

CRISP - und die Gliederung des Kurses

Strukturierte Daten

Aufbau und Organisation eines Data Frame

In [1]:

import pandas as pd
data = pd.read_csv('../3_data/bank_data.csv', sep=';')
#print(data.iloc[0:11, 0:6])

## arbitrarily input
data.iloc[3, 0] = None
data.iloc[6, 1] = None
#print(data.iloc[0:11, 0:6])
print(data.iloc[0:11, [0,1,2,10,11,20]])

     age            job  marital  duration  campaign    y
0   30.0  self-employed   single     245.0         3  yes
1   32.0     technician  married     370.0         1   no
2   27.0    blue collar   single     623.0         1  yes
3    NaN    blue collar   single       9.0         6   no
4   27.0         admin.   single     126.0         2  yes
5   34.0         admin.   single     548.0         2  yes
6   46.0           None  married      86.0         2   no
7    NaN        retired  married     707.0         3  yes
8   46.0         admin.  married      96.0         6   no
9   48.0    blue collar  married     241.0         2   no
10  29.0     technician  married     154.0         3   no

Begriffe

Beispieldaten

Cell:

[Cell type raw - unsupported, skipped]

Cell:

[Cell type raw - unsupported, skipped]

Anforderungen an die Daten für Machine Learning

Eine typische ML Sequenz

Python Libraries

Feature Engineering

Die Python Libraries und CRISP-DM

Begleitende Literatur