Go to TogaWare.com Home Page. Data Science Desktop Survival Guide
by Graham Williams
Duck Duck Go

Feature Selection

20180726 The FSelector package provides functions to identify subsets of variables that might be more effective for modelling. We can use this (and other packages) to assist us in reducing the variables that will be useful in our modelling. As we find useful functionality we will add them to our standard template so that for our next dataset we have the functionality readily available.

We first use FSelector::cfs() to identify a good subset of variables using correlation and entropy. We then list the variable importance using FSelector::information.gain() to advise a useful subset of variables. Note that the stringi::https://www.rdocumentation.org/packages/stringi/topics/to concatenate strings together to produce a formula that indicates we will model the target variable using all of the other variables of the dataset.

# Construct the formulation of the modelling we plan to do.

form <- formula(target %s+% " ~ .") %T>% print()
## rain_tomorrow ~ .

# Use correlation search to identify key variables.

cfs(form, ds[vars])
## [1] "rainfall"     "sunshine"     "humidity_3pm" "cloud_3pm"    "rain_to...

# Use information gain to identify variable importance.

information.gain(form, ds[vars])
##                 attr_importance
## min_temp            0.005574465
## max_temp            0.014306291
## rainfall            0.058146788
## evaporation         0.004816268
## sunshine            0.050838588
## wind_gust_dir       0.005941322
## wind_gust_speed     0.026772843
## wind_dir_9am        0.008740164
## wind_dir_3pm        0.004717540
## wind_speed_9am      0.004189375
## wind_speed_3pm      0.005426695
## humidity_9am        0.038093685
## humidity_3pm        0.108910707
## pressure_9am        0.027130527
## cloud_9am           0.032297858
## cloud_3pm           0.047524832
## rain_today          0.046474217

The two measures are consistent in this case in that the variables identified by FSelector::cfs() are the more important variables identified by FSelector::information.gain().


Support further development by purchasing the PDF version of the book.
Other online resources include the GNU/Linux Desktop Survival Guide.
Books available on Amazon include Data Mining with Rattle and Essentials of Data Science.
Popular open source software includes rattle and wajig.
Hosted by Togaware, a pioneer of free and open source software since 1984.
Copyright © 2000-2020 Togaware Pty Ltd. . Creative Commons ShareAlike V4.