Skip to content

Background

NeuroPySeminar

Advances in Data Analysis: Python

An interactive Python-based seminar repository diving into contemporary data analysis methods from recent research papers. Engage hands-on with real data, explore foundational theories, and focus on techniques in time series analysis, dimensionality reduction, and dynamical systems.

Methods


W01
EMD
TimeSeries
Link
W02
Multitaper
TimeSeries
Link
W03
AR
TimeSeries
Link
Extra
W04
PCA
DimReduction
Link

W05
ICA
DimReduction
Link
W06
UMAP
DimReduction
Link
W07
CCA
DimReduction
Link
Extra
W08
NNMF
DimReduction
Link

W09
GPFA
Dynamics
Link
W10
SINDy
Dynamics
Link
W11
Network
Dynamics
Link
W12
HMM
Dynamics
Link

GitHub: NeuroPySeminar

GitHub-Pages: NeuroPySeminar

PaperPool

Slides

ZoomRecordings

LSF

Moodle

Course Description

Type of Course: Seminar

LSF Number: 19409

Term: WiSe2025/26

Max. participants: 12

Language: English

Title: Advances in Data Analysis: Python

Shahidi, Arash, Phd Student at Sirota Lab

Faculty of Biology - Ludwig-Maximilian University of Munich

This seminar introduces recent data analysis methods highlighted in current research papers. Using Python, participants will implement these techniques and apply to data, either publicly available or data from their own projects. We'll particularly focus on analysis methods regarding time series data, dimensionality reduction, and dynamical systems. The foundational theories behind these methods will be discussed, referencing established analytical texts.

Prerequisites

🌱 Minimal Prerequisites for Learning the Methods

  1. Python Basics 🐍

    • Write simple scripts.

    • Use libraries like numpy, matplotlib, and pandas.

  2. Linear Algebra (lightweight) ➕

    • Vectors and matrices.

    • Matrix multiplication, eigenvalues/eigenvectors.

  3. Probability & Statistics (essentials) 🎲

    • Mean, variance, correlation.

    • Gaussian (normal) distribution.

  4. Signal Processing & Fourier Basics 🎶

    • What is frequency?

    • Fourier transform for breaking signals into components.

  5. Machine Learning Concepts (intro level) 🤖

    • What “dimensionality reduction” means (compressing data).

    • What “clustering” means (grouping similar things).

✨ That’s enough to start exploring PCA, ICA, EMD, UMAP, and the rest. You don’t need to be an expert — just some curiosity and willingness to learn will take you a long way.

Assessment

  • 60% Main Presentations: Presenting (at least) one theory session (50 min) along with the corresponding exercise session (30 min).
    • send your top 3 choices from the paper pool to the instructor.
  • 40% Short Exercise Presentations: Presenting 3 exercises (10 min /exercise) on 3 topics other than their selected topic.
    • Other than the provided exercises, applying the methods to any publicly available data, simulated demos, or data from personal projects, also counts as an exercise and is encouraged.

Text Books

In addition to the papers, the following books will be referred to

  • Observed Brain Dynamics, Mitra & Bokil
  • Advanced Data Analysis in Neuroscience, Daniel Durstewitz

  • https://goodresearch.dev/ A short Handbook on how to setup and organize your projects in Python.
  • VSCode: A popular IDE with an abundance of plugins that make coding easier
  • Github Copilot : AI coder added as a plugin to VSCode - Free for all students and teachers Apply for GitHub Education Benefits.
  • Google Colab

Other useful tools: - Obsidian - Zotero - ChatGPT, Claude, etc.