Chapter 8, Clustering and Classification, involves quintessential analysis methods involving k-means clustering, hierarchical clustering, constructing decision trees, and implementing the k-Nearest Neighbors classifier.
This is the accompanying source code for Haskell Data Analysis Cookbook. Refer to the book for step-by-step explanations.
- Code01: Implementing the k-means clustering algorithm
- Code02: Implementing hierarchical clustering
- Code03: Using a hierarchical clustering library
- Code04: Finding the number of clusters
- Code05: Clustering words by their lexemes
- Code06: Classifying the parts of speech of words
- Code07: Identifying key words in a corpus of text
- Code08: Training a parts of speech tagger
- Code09: Implementing a decision tree classifier
- Code10: Implementing a k-Nearest Neighbors classifier
- Code11: Visualizing points using Graphics.EasyPlot
Install the Haskell Platform.
$ sudo apt-get install haskell-platform
Alternatively, install GHC 7.6 (or above) and Cabal.
$ sudo apt-get install ghc cabal-install
A Makefile
is provided in each recipe. Compile the corresponding executable by running make
.
$ make
Run the resulting code. For example,
$ ./Code01
To clean up the directory:
$ make clean