Mercurial > repos > bgruening > sklearn_feature_selection

annotate README.rst @ 0:2bbacfaadb5c draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
author bgruening
date Thu, 12 Apr 2018 08:19:35 -0400
parents
children 0cd9e541bf7c
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
rev   line source
0
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
1 ***************
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
2 Galaxy wrapper for scikit-learn library
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
3 ***************
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
4
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
5 Contents
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
6 ========
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
7 - `What is scikit-learn?`_
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
8 - `Scikit-learn main package groups`_
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
9 - `Tools offered by this wrapper`_
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
10
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
11 - `Machine learning workflows`_
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
12 - `Supervised learning workflows`_
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
13 - `Unsupervised learning workflows`_
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
14
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
15
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
16 ____________________________
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
17
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
18
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
19 .. _What is scikit-learn?
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
20
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
21 What is scikit-learn?
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
22 ===========================
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
23
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
24 Scikit-learn is an open-source machine learning library for the Python programming language. It offers various algorithms for performing supervised and unsupervised learning as well as data preprocessing and transformation, model selection and evaluation, and dataset utilities. It is built upon SciPy (Scientific Python) library.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
25
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
26 Scikit-learn source code can be accessed at https://github.com/scikit-learn/scikit-learn.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
27 Detailed installation instructions can be found at http://scikit-learn.org/stable/install.html
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
28
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
29
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
30 .. _Scikit-learn main package groups:
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
31
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
32 ======
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
33 Scikit-learn main package groups
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
34 ======
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
35
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
36 Scikit-learn provides the users with several main groups of related operations.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
37 These are:
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
38
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
39 - Classification
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
40 - Identifying to which category an object belongs.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
41 - Regression
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
42 - Predicting a continuous-valued attribute associated with an object.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
43 - Clustering
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
44 - Automatic grouping of similar objects into sets.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
45 - Preprocessing
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
46 - Feature extraction and normalization.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
47 - Model selection and evaluation
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
48 - Comparing, validating and choosing parameters and models.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
49 - Dimensionality reduction
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
50 - Reducing the number of random variables to consider.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
51
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
52 Each group consists of a number of well-known algorithms from the category. For example, one can find hierarchical, spectral, kmeans, and other clustering methods in sklearn.cluster package.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
53
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
54
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
55 .. _Tools offered by this wrapper:
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
56
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
57 ===================
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
58 Available tools in the current wrapper
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
59 ===================
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
60
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
61 The current release of the wrapper offers a subset of the packages from scikit-learn library. You can find:
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
62
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
63 - A subset of classification metric functions
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
64 - Linear and quadratic discriminant classifiers
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
65 - Random forest and Ada boost classifiers and regressors
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
66 - All the clustering methods
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
67 - All support vector machine classifiers
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
68 - A subset of data preprocessing estimator classes
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
69 - Pairwise metric measurement functions
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
70
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
71 In addition, several tools for performing matrix operations, generating problem-specific datasets, and encoding text and extracting features have been prepared to help the user with more advanced operations.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
72
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
73 .. _Machine learning workflows:
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
74
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
75 Machine learning workflows
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
76 ===============
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
77
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
78 Machine learning is about processes. No matter what machine learning algorithm we use, we can apply typical workflows and dataflows to produce more robust models and better predictions.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
79 Here we discuss supervised and unsupervised learning workflows.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
80
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
81 .. _Supervised learning workflows:
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
82
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
83 ===================
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
84 Supervised machine learning workflows
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
85 ===================
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
86
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
87 **What is supervised learning?**
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
88
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
89 In this machine learning task, given sample data which are labeled, the aim is to build a model which can predict the labels for new observations.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
90 In practice, there are five steps which we can go through to start from raw input data and end up getting reasonable predictions for new samples:
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
91
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
92 1. Preprocess the data::
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
93
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
94 * Change the collected data into the proper format and datatype.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
95 * Adjust the data quality by filling the missing values, performing
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
96 required scaling and normalizations, etc.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
97 * Extract features which are the most meaningfull for the learning task.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
98 * Split the ready dataset into training and test samples.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
99
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
100 2. Choose an algorithm::
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
101
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
102 * These factors help one to choose a learning algorithm:
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
103 - Nature of the data (e.g. linear vs. nonlinear data)
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
104 - Structure of the predicted output (e.g. binary vs. multilabel classification)
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
105 - Memory and time usage of the training
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
106 - Predictive accuracy on new data
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
107 - Interpretability of the predictions
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
108
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
109 3. Choose a validation method
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
110
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
111 Every machine learning model should be evaluated before being put into practicical use.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
112 There are numerous performance metrics to evaluate machine learning models.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
113 For supervised learning, usually classification or regression metrics are used.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
114
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
115 A validation method helps to evaluate the performance metrics of a trained model in order
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
116 to optimize its performance or ultimately switch to a more efficient model.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
117 Cross-validation is a known validation method.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
118
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
119 4. Fit a model
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
120
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
121 Given the learning algorithm, validation method, and performance metric(s)
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
122 repeat the following steps::
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
123
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
124 * Train the model.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
125 * Evaluate based on metrics.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
126 * Optimize unitl satisfied.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
127
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
128 5. Use fitted model for prediction::
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
129
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
130 This is a final evaluation in which, the optimized model is used to make predictions
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
131 on unseen (here test) samples. After this, the model is put into production.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
132
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
133 .. _Unsupervised learning workflows:
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
134
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
135 =======================
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
136 Unsupervised machine learning workflows
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
137 =======================
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
138
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
139 **What is unsupervised learning?**
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
140
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
141 Unlike supervised learning and more liklely in real life, here the initial data is not labeled.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
142 The task is to extract the structure from the data and group the samples based on their similarities.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
143 Clustering and dimensionality reduction are two famous examples of unsupervised learning tasks.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
144
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
145 In this case, the workflow is as follows::
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
146
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
147 * Preprocess the data (without splitting to train and test).
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
148 * Train a model.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
149 * Evaluate and tune parameters.
2bbacfaadb5c planemo upload for repository https://github.com/bgruening/galaxytools/tools/sklearn commit 7a31960686122d7e53054fef4996525f04ebd254
bgruening
parents:
diff changeset
150 * Analyse the model and test on real data.