Machine learning tutorials for NSCI0028
In this tutorial, we introduce first the basics to start using Python for data mining and machine learning.
The student should follow the following steps:
1- Start by copying the link to the Jupyter Notebook from the folder in GitHub entitled "Basics_in_Python.ipynb"
2- Open Colab in the link below: https://colab.research.google.com/
3- Select the tab Github and paste the link in the dedicated field to search for the Notebook
4- Open the link of the Notebook and press Connect
5- Each time, you would like to run an exercise, press on the small triangle in the black circle and wait for a few second to see the output underneath the code.
6- When you finish all the exercice, open a fresh colab session and choose another notebook.
7- Follow this order of the notebooks:
a- Basics_in_Python.ipynb
b-Introduction_to_NumPy.ipynb
c-Introduction_to_Pandas.ipynb
d-Basics_visualisation.ipynb
e-Loops and functions.ipynb
f-Preprocessing.ipynb
g-Preprocessing_Iris_dataset.ipynb
h-Feature_selection.ipynb.ipynb
i-Underfitting_Overfitting.ipynb
j-Outliers_removal.ipynb
k-Multiple_Linear_Regression.ipynb
l-Random_Forest_Iris.ipynb
m-KNN.ipynb
n-KNN-optimsation.ipynb
o-SVM.ipynb
p-imbalanced_classes.ipynb
q-Kmeans.ipynb
r-kmeans_2.ipynb
s-hierarchical_clustering.ipynb
t-PCA_1.ipynb
u-PCA_2.ipynb
v-PCA_3.ipynb
w-Logistic_regression.ipynb
Install and run Knime
- Create a directory to install it in eg:
NOTE: your ssh connection to Aristotle will need X tunnelling enabled. So, run Xming or Exceed then run Putty and activate X11.
I. Installation of Knime
- Run the following in the terminal:
mkdir -p ~/apps/KNIME
cd ~/apps/KNIME
- Download the 64bit Linux version, unarchive it and update your PATH:
wget https://download.knime.org/analytics-platform/linux/knime-latest-linux.gtk.x86_64.tar.gz
tar -xvf knime-latest-linux.gtk.x86_64.tar.gz
export PATH=$PATH:~/apps/KNIME/knime_4.7.1
- Load the following modules:
module -f unload compilers mpi gcc-libs
module load beta-modules
module load gcc-libs/10.2.0
II- Running Knime:
- Run the following in the terminal:
cd ~/apps/KNIME
export PATH=$PATH:~/apps/KNIME/knime_4.7.1
module -f unload compilers mpi gcc-libs
module load beta-modules
module load gcc-libs/10.2.0
knime &