Spatial Data Science

The module Spatial Data Science introduces the basics of stochastics and descriptive statistics, and advancing through inferential statistics, machine learning, and neural networks. With practical applications in mind, we explore how probability, hypothesis testing, and statistical analysis form the backbone of data science, especially in the spatial context. The focus is on practical application, as well as understanding the principle of methods and their limitations.

As we progress into machine learning, we examine different types of problems such as classification, regression, and clustering, in which we gain practical experience in handling tabular data. The challenge of overfitting and underfitting will be explored, along with methods to evaluate and prevent it. Moving into neural networks, we cover basic principles and apply various architectures, from simple feed-forward networks to convolutional and residual networks, which are especially valuable for analyzing spatial data like satellite imagery. Advanced concepts, such as hyperparameter optimization and fine-tuning models help to increase the performance of deep learning models.

10 Feb
February 10, 2025 - May 12, 2025
Spatial Data Science – Registration
02 Jun
June 2, 2025 - September 1, 2025
Spatial Data Science – Registration
06 Oct
October 6, 2025 - January 5, 2026
Spatial Data Science – Registration

3 months

English

6 ECTS

The module is free of charge for UNIGIS students working to meet their elective subject requirements. Included are:

  • all related study materials
  • supervision and assessment
  • module accreditation according to the curriculum

ClubUNIGIS members can register at a price of € 350,-. Included are:

  • all related study materials
  • supervision and assessment
  • course certificate upon completion of the module

Data Science has gained significant attention over the past few decades due to its successes in various areas, ranging from analyzing satellite images to predicting economic trends. This module aims to provide a clear and practical overview of Data Science, focusing on its applications in the spatial context. We will cut through the hype and offer concrete examples of how these concepts are used in practice. To start, we’ll define some of the most important terminology in the field.

The module Spatial Data Science requires more existing expertise than other elective modules. The level of the module is above average, as mastering mathematics skills at A-level and some experience with the Python programming language are ultimately expected. As Spatial Data Science encompasses a wide range of topics, from data visualization and inferential statistics to databases, algorithms, and advanced concepts in artificial intelligence, some chapters might be challenging. Additional resources for further study will be provided for certain topics.

If you are not sure whether you qualify, please contact martin.loidl@plus.ac.at.

This is an exclusively english language module, hence any kind of communication with the module lecturer should be in English. A discussion forum is maintained in Moodle in order to support efficient module instruction. You are requested to submit all your questions related to this module to this forum only. The lecturer will check all incoming comments on a regular basis. Theoretical concepts are complemented with practice oriented examples demonstrated with help of multimedia elements. Upon completion of the module students are requested to evaluate the module, which is a part of our quality assurance policy and practice.

Python, sklearn, PyTorch, Google Colab / Jupyter

Lecturer assessment reflects student’s achievements in this module and is conducted through assessing module assignments. Assignments (English language!) must be submitted in written format (.PDF) within the required time period. Each coding example (optionally also the others for consistency) must be written in a Jupyter notebook, the ipynb file can be exported as html or pdf. Exercises are designed to enforce students’ knowledge and skills. These should be completed to allow students to assess their own progress and are not included in the module assessment.

Lesson 1 – Just Explain It in a Nutshell

Definition of the most important terminology in the field.

Lesson 2 – Does AI Really Run On Bare Metal?

Topics such as circuit boards, digital arithmetic, compilers, and low-level instruction are beyond the scope of this course. Nonetheless, you will set up a minimal toolstack that allows for GPU-accelerated calculations.

Lesson 3 – The Foundation of All Is: Stochastics

Basics of stochastics and understanding distributions

Lesson 4 – Explaining Data Using: Descriptive Statistics

Understanding variables,  Descriptive Statistics, Understanding Conditional Probability

Lesson 5 – The Pattern Behind Samples: Distributions

Understanding that statistic is dealing with unknown distributions.

Lesson 6 – But We Don’t Know the Real Parameters: Inferential Statistics

Being able to interpret and employ statistical tests.

Lesson 7 – A Brief Overview On: Machine Learning

Get a good overview over the field of Machine Learning. Understand what Machine Learning methods do. Being able to apply several ML-methods on tabular data. Understanding Overfitting

Lesson 8 – How Can Machines Learn To: Classify

Basics of classification in machine learning and classification models.

Lesson 9 – Concepts You Need to Know About Machine Learning

This lesson gives an overview of concepts essential for classification and machine learning in general.

Lesson 10 – How Machines Estimate Based on Data?

Key concepts and methods in machine learning.

Lesson 11 – Now We Come To: Artificial Neural Networks

This lesson helps you to understand the basic structure of an artificial neural network and what the parameters do.

Lesson 12 – Here Is a Drone Image, Let’s Classify It

In this lesson, we consider more complex input data, such as image data originating from drones and satellites.

Lesson 13 – Getting Deeper on AI

In this lesson, we’ll explore some of the most effective and widely used methods in deep learning, particularly for analyzing complex datasets like satellite images.

Lesson 14 – Is There More on Data Science?

This lesson complements other commonly used topics in data science.

Lesson 15 – Outlook to Your Own Learning Experience

This lesson contains resources and additional information to continue building on your foundation.