Introduction to Machine learning

This is the main repository of the Machine learning course teach at November/December of 2017. You can find different folder containing the dataset and content needed for the course.

Table of content

• Slides
• Getting Started
  • Pre Setup
  • Python Installation
  • Environment Setup
• Interesting links
• Questions
• Previous courses

Slides

Class	Slides	Solutions
Session 1 - The What Why and when of Machine Learning	Slides	Solutions
Session 2 - Feature Engineering	Slides	Solutions
Session 3 - Linear Classifiers	Slides
Session 4 - Classifier Optimization	Slides
Session 5 - Neural Network Overview	Slides

Getting Started

Pre Setup

• Make an account on github.com

  • Does not have to be your regular account. You could also make a temporary one just for this class/project
  • You will use this account for all your projects
  • You will have the option to set up a landing web page to publish your project beautifully as well

• Send us your github ID

  • This is very important, as we will be assigning the teams according to your IDs
  • Also send us your knowledge level in Machine learning (Beginner/Intermediate/Expert)

• If you are new to git and/or github and will be using it from windows

  • Download GitHub Desktop (https://desktop.github.com/ )
  • Go through a couple of simple git tutorials understanding the purpose of git (https://guides.github.com/activities/hello-world/)

• If you're familiar with python, you can skip the Python setup section

  • If not, I'd suggest you install python the same way mentioned

Python Installation

Windows Installation

• For windows, the easiest thing to do is to install a distribution of python, and not just the raw python installation.

• One of the most popular distributions of Python is Anaconda (https://www.anaconda.com/download/)

• Download the 3.6 version (because more future compatibility)

  • For our purposes, we don't care as long as version is greater that 2.7

• Follow Installation instructions

  • Add it to your path, as that would make life so much easier.
  • Do it during the installation
  • or if you know what you're doing, set it later. [Not Recommended]

Linux Installation

• Assuming that you're an expert
  • sudo apt-get install python3.6
  • sudo apt-get install jupyter-notebook python-scipy python-spyder

Environment Setup

Anaconda (Windows)

• Open up Anaconda Navigator
• If it asks you to make a virtual environment, do that with the default settings
• Open up an anaconda prompt
• install the required packages conda install seaborn scikit-learn matplotlib *// [Optional]// if you like a scientific IDE - conda install spyder

Linux

• if you do not have a virtual-env : sudo pip install matplotlib seaborn scikit-learn scipy numpy notebook

• if you do have a virtual env, skip the sudo

• Cheers! You're ready to go! Open up spyder or notebook

Cython==0.26.1
ipykernel==4.6.1
ipython==6.1.0
ipython-genutils==0.2.0
ipywidgets==7.0.0
jupyter-client==5.1.0
jupyter-console==5.2.0
jupyter-core==4.3.0
jupyterlab==0.27.0
jupyterlab-launcher==0.4.0
matplotlib==2.0.2
notebook==5.0.0
numpy==1.13.1
numpydoc==0.7.0
pandas==0.20.3
pandocfilters==1.4.2
scikit-image==0.13.0
scikit-learn==0.19.0
scipy==0.19.1
seaborn==0.8
tensorflow==1.4.0
tensorflow-tensorboard==0.4.0rc3

Interesting links

Session 1

• Common myths in ML - Good article.
• Matplotlib cheatsheet - Cheatsheet for using Python's matplotlib library.
• Pandas cheatsheet - Cheatsheet for using Python's pandas library.
• Pandas cookbook - Recipes for using Python's pandas library
• Seaborn cheatsheet - Cheatsheet for using Python's seaborn library.
• Two Minute Papers - Youtube channel with interesting ML videos.
• Violin Plots - Description and examples of use.

Session 2

• Dataset - Differences between training, test and validation dataset.
• Feature engineering Tips on feature engineering
• Feature selection - Introduction to feature selection methods.
• Pearson's correlation - Simple explanation of the pearson's correlation method.
• PCA - PCA Algoritm info
• PCA2 - More PCA algorithm info.

Session 3

• Random Forest - Good video for understand the algorithm.
• Random Forest 2 - Random forest algorithm introduction with explanations of the main concepts.
• SVM - SVM algorithm simple explanation.
• SVM 2 - SVM algorithm advanced explanation.
• SVM 3 - Demo.
• Visual Machine Learnign - A visual introduction to machine learning.

Session 4

• Ensamble methods - Description of the ensambled methods.
• Sklearn - Examples of use of this tool.

Questions

If you want to ask something, feel free to write your question in the issues section.

Courses

• November '17 (12 assistants)
• January '18