Tabular data

fast.ai reference

fast.ai course material book

Search…

Fast.ai reference

Fully connected layer

NLP

Reinforcement Learning

Overfitting and Underfitting

Contribute

FEEDBACK & CORRECTION

Tabular data

Tabular data is data that is in tables. This is one of the most prevalent forms of data in industry, yet underrepresented in research. Things like information about real estate, bank accounts, user profiles, spreadsheets etc... are usually stored in tables.
site_name
state
elevation
MMI
EPT_PTAX
data_source
agency
Flat Creek
OR
1571
26.31
21.62
WSA
EPA
Barlow Creek
OR
1010
79.12
53.49
WSA
EPA
Stevens Creek
WA
1051
61.31
61.54
WSA
EPA
Fast.ai features a very handy library to create and train models based on tabular data, import it as follows:
from fastai.tabular import *
Categorical and Continuous data
Tabular data usually consists of two different kinds of data. Categorical data is represents data that can be divided into unique groups. Food types, gender, marital status etc...
Continuous data is data which exists at all points of an interval. For example, the price of a stock:
The stock price fluctuates continuously
These two types of data have to be treated differently. If you were to treat food type as a continuous datatype, then your model might classify your red strawberry ice-cream image as an ice-steak.
The fast.ai library makes this easy:
cat_names = ['workclass', 'education', 'marital-status',  'relationship', 'race']
cont_names = ['age', 'salary', 'square-footage-house']
​
# and then you just indicate the datatypes to the TabularList function 
data = TabularList.from_df(cat_names=cat_names, cont_names=cont_names)
 The data object can then be fed directly into your model.

Pooling

Fully connected layer

Last modified 2yr ago

Export as PDF

Copy link