ML Classroom

The education and support of young researchers is an important part of the work at the Lamarr Institute. Many of our scientists are therefore involved in teaching at our partner institutions, passing on their expertise to the next generation. As part of a seminar on science communication, computer science students learn how to prepare Machine Learning topics for a broader public audience. In this blog series, the Lamarr Institute offers these emerging researchers a platform to showcase their newly acquired skills and publish their contributions.

Basics

Clustering for recognizing medical patterns: Gaussian Mixture Models explained

24. April 2024

Medical data often hides patterns that are difficult to recognize but relevant for diagnostics and therapy. How can we identify these structures? The clustering algorithm Gaussian Mixture Model provides an answer....

00 Blog Classroom Beitrag2 Pruningalgorithmen 2 FraundWingenfel - Lamarr Institute for Machine Learning (ML) and Artificial Intelligence (AI)

Basics

Two Pruning Algorithms: MEP vs. PEP – one Goal, different Outcomes

28. February 2024

Pruning can simplify complex decision trees. In this comparison , we explore two algorithms, Minimum Error Pruning (MEP) and Pessimistic Error Pruning (PEP). What are the differences between them, and how far should one go with pruning?...

00 Blog Classroom Beitrag1 Decision Trees FraundWingenfel - Lamarr Institute for Machine Learning (ML) and Artificial Intelligence (AI)

Basics

Complex Decision Trees? How Pruning keeps the ML Tool organized

14. February 2024

Decision Trees are one of the most easily understandable and straightforward tools in the field of Machine Learning. However, Big Data has made them more complex in recent times. Can pruning help make Decision Trees accessible once again?...

00 Blog Studis fur Studis GraphSAGE - Lamarr Institute for Machine Learning (ML) and Artificial Intelligence (AI)

Basics

GraphSAGE: Predicting user behavior with ML-based recommendation services

23. May 2023

GraphSAGE is a powerful ML method that is used to analyze large-scale and dynamically growing data with a graph structure. How exactly does the method, which can be used to provide personalized recommendations, work?...

00 Blog Studis fuer Studis CSC 1 - Lamarr Institute for Machine Learning (ML) and Artificial Intelligence (AI)

Basics

Automatic classification of a publication network using a Graph Attention Network

19. May 2023

Social networks are omnipresent these days. But how can information be extracted from them with the help of neural networks? To answer this question, we present the Graph Attention Network....

00 Blog Studis fuer Studis CSC - Lamarr Institute for Machine Learning (ML) and Artificial Intelligence (AI)

Basics

Graph Neural Networks: An introduction to the world of graph-based AI models

10. May 2023

Numerous everyday structures can be represented as graphs. Graph neural networks use their structure to collect information from neighboring nodes and to analyze the relationships between objects in graphs....

00 Blog Studis fuer Studis LDA - Lamarr Institute for Machine Learning (ML) and Artificial Intelligence (AI)

Basics

Linear discriminant analysis: A method for dimension reduction of classifiable data sets

19. April 2023

Large amounts of data are problematic when calculating ML algorithms. One solution lies in dimension reduction methods such as linear discriminant analysis (LDA), which reduces the number of data features while taking into account the existing class labels....

00 Blog Studis fuer Studis PCA - Lamarr Institute for Machine Learning (ML) and Artificial Intelligence (AI)

Basics

Principal component analysis – A method for dimension reduction on large amounts of data

12. April 2023

Learning on large, high-dimensional data sets provides particularly meaningful results, but consumes a lot of memory and computing capacity. The dimension reduction method can help to solve this problem....