Repository files navigation

Notebooks and Scripts on Machine Learning

1. Introduction
2. Supervised Learning link
   1. Introductionlink
   2. Regression
      1. Introduction link
      2. Linear Regressionlink
      3. Support Vector Machine (SVM)
      4. Lasso
      5. Ridge
   3. Classificationlink
      1. Introduction link
      2. Support Vector Machine (SVM)link
      3. Principal Component Analysis (PCA)link
      4. Kernel Principal Component Analysis (KPCA)link
3. Unsupervised Learning link

This notebooks just shows some examples of the methods collected from tutorials, books and ourselves in order to get some insights on the ideas of supervised and unsupervised learning.

This is only for learning purposes.

There are a lot of resources in this respect and this notebooks do not pretend to be a complete and detailed description, just some vague idea of which kind of problems may be addressed with which strategy.

Visualization is always important when working with data, nevertheless it is not always possible as there are high dimensional dataset that makes this task a complete challenge, so to illustrate here we are going to use 2D and 3D data and different libraries introduced when needed.

The most difficult part during the process of working with data & ML, is getting clean the dataset, so we will face those problems carefully.

Some people describe Machine learning in comparison withnormal programming, as follows

When programming, normally you have some inputs and some rules, and the code computes an outputs, but on machine learning, you give the inputs and outputs and the computer gives you therule

I particularly think that there is something missing in that description and is that you must know something about the phenomena, so that amodel is also given, and the output is basically the best way our model describes the data we gave, namely the input and output. So therule we get is based on our model.

So the question is always how to choose our model.

Here you can find some links of courses, tutorials and notebooks.

• Data Sciencelink: Machine learning itself is not enough. Data analysis is required so that machine learning makes sense, here is one well done, detailed and intereactivecourse on data science.

• Notes on Machine Learninglink: This is a study material very detailed and easy to follow.

• Machine Learning Materiallink: A Curated list on machine learning material.Awesome ML

• Machine Learning Booklink: Repository of the bookpython Machine Learning

This list will be updated on the following days.

My name is Mauricio Sevilla, a physicist who enjoys to program and teach so this is the case for both of them.

emailjmasevillam@gmail.com