とらりもんHOME  Index  Search  Changes  Login

とらりもん - Python training course - exercise 2 Diff

  • Added parts are displayed like this.
  • Deleted parts are displayed like this.

2018/05/27 Ishibashi

[[link to Python training course|http://pen.agbi.tsukuba.ac.jp/~ryuiki_plus/hiki/?Python+training]] course|http://pen.envr.tsukuba.ac.jp/~ryuiki_plus/hiki/?Python+training]] in hiki (only lab network.)

[[link to exercise 1|http://pen.agbi.tsukuba.ac.jp/~torarimon/?Python+training+course+-+exercise+1]]1|http://pen.envr.tsukuba.ac.jp/~torarimon/?Python+training+course+-+exercise+1]]

[[link to answer|http://pen.envr.tsukuba.ac.jp/~torarimon/?exercise+2+answer]]

! Exercise 2
In exercise 2, we will additionally use scikit-learn.

Through this exercise, you can be able to handle image data using Python, how to use Python machine learning libraries and how to conduct simple accuracy assessment.

A learner is expected to finish [["Learn the Basics" & "Data Science Tutorials"|https://www.learnpython.org/]] and [[exercise 1|http://pen.agbi.tsukuba.ac.jp/~torarimon/?Python+training+course+-+exercise+1]] 1|http://pen.envr.tsukuba.ac.jp/~torarimon/?Python+training+course+-+exercise+1]] before starting this exercise.

!! Exercise 2: Machine learning and validation.
we will use MNIST dataset. please search, if you do not know what it is.
* Exercise 2.1: import MINIST dataset into Python using scikit-learn library.

* Exercise 2.2: see the data number and pixel size of imported data.

* Exercise 2.3: save one or two imported image in Python into your PC.

* Exercise 2.4: conduct machine learning with scikit-learn. (You can
decrease the number of training data, if the running time is too long.)

* Exercise 2.5: calculate overall accuracy.