Skip to content

michalgregor/ml_class

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Educational Material, Machine Learning

This repository holds and links to educational material related to machine learning; most of it is from my machine learning class at the University of Žilina. Some material is from other classes, where I participated.

License

The content is made available under the CC-BY license.

Disclaimer

Note that the content is being shared as is; it may contain errors, outdated information + I am not keeping the code of the notebooks up-to-date at the moment, so they are liable to be broken by package updates. So consider yourselves warned.

Machine Learning

PowerPoint Slides

The PowerPoint slides for my machine learning class are hosted in this folder. The slides are in English with Slovak translations provided in the slide notes. The following topics are covered:

  • ML1: Introduction and Key Concepts;
  • ML2: Simple Machine Learning Methods;
  • ML3.1: Data Analysis;
  • ML3.2: Cluster Analysis;
  • ML4: Supervised Learning and Optimization;
  • ML5: Evaluation, Regularization, Interpretation;
  • ML6: Artificial Neural Networks and Automatic Differentiation;
  • ML7: Deep Learning;
  • ML8: Deep Learning for Sequential Data;
  • ML9.1: Dimensionality Reduction;
  • ML9.2: Embeddings and Face Clustering;
  • ML10: Reinforcement Learning, Value-Based Methods;
    • ML10: Reinforcement Learning (1-lecture version);
  • ML11: Deep Reinforcement Learning;
  • ML12: Gaussian Processes and Bayesian Hyperparameter Optimization;

Colab Notebooks

The Colab notebooks for the class are here (note that these are the student versions; teacher versions are below):

Video Lectures

Computer Vision

Note: This only covers the part of the course that I taught, which was the deep learning part.

PowerPoint Slides

The PowerPoint slides are hosted in this folder. Again, the slides are in English with Slovak translations provided in the slide notes. The following topics are covered:

  • CV1: Recap of deep learning + the specific tasks and challenges of computer vision;
  • CV2: Visual object detection using deep learning;
  • CV3: Semantic segmentation using deep learning;
  • CV4: Few-Shot Learning of Fine-Grained Visual Concepts;

Video Lectures

Data Analysis

Note: This only covers the part of the course that I taught, which was machine-learning-oriented – not the part on probability and statistics.

PowerPoint Slides

The PowerPoint slides are hosted in this folder. Again, the slides are in English with Slovak translations provided in the slide notes. The following topics are covered:

  • DA1: Data Analysis: introduction + a segment on exploratory data analysis

  • DA2: Intro to Modelling;

    • Basics of machine learning;
    • Decision trees;
    • Linear regression (prediction PoV);
    • Logistic regression (prediction PoV);
    • Ensemble methods;
  • DA3: Intro to Preprocessing;

    • Categorical data;
    • Numeric data;
    • Advanced preprocessing methods;
    • Learning to preprocess unstructured data;
    • Missing data;
  • DA4: Verification;

    • Types of validation;
    • Stratification;
    • Performance indicators (for classification and regression);
    • Underfitting, overfitting and regularization;
  • DA5: Statistical Inference using Linear Regression;

    • Linear regression (statistical inference PoV);
    • Interpretation of results;
    • Model selection;

Colab Notebooks

The Colab notebooks for the class are here (note that these are the student versions; teacher versions are below):

Video Lectures

Colab Notebooks' Sources + Solutions (For Teachers)

Colab notebooks' sources are hosted jointly in a single repository, which is only accessible upon request to teachers, because it contains solutions. Most Colab notebooks use cell tagging that allows multiple different versions to be exported automatically: this includes exporting student and teacher versions as well as English and Slovak versions. The tagging system allows efficient joint maintenance of all versions in the same file.

You can request access through THIS FORM. The code is shared under a permissive license, but with the request not to share the teacher versions publicly.

About

Repository that links to material for a machine learning class.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published