LINKING DATA MINING, SPATIAL ANALYSIS AND ALGORITHMIC DESIGN THROUGH PYTHON

 

ABOUT

The aim of the workshop, is to introduce participants to the data mining and machine learning fields and their use in spatial analysis and design. Striping these subjects of their inherent complexity participates will explore them by playing with tools and code, focusing on simple tasks in a hands-on approach. Attendants will have an introduction to Python programming and its power as a scripting language in tools of diversified fields beyond strict data analysis (space syntax: Depthmap; algorithmic design: Grasshopper; GIS: QGIS; etc), but some data mining visual programming interfaces will also be demonstrated. We will focus on the popular Python data science benchwork Anaconda, through Jupyter Notebooks, that turn Python coding into a playful, interactive and shareable process. In the end attendants should get an overview on the data mining field and Python scripting, and recognise the potential of its use in their own work or investigation and where to look for further information .

 

PROGRAMME

The workshop of 8 hours includes the following topics:

  • Brief introduction to data, data mining, machine learning and related tools.
  • Introduction to Anaconda Python Data Science Platform and Jupyter Notebook.
  • Crash course on Python.
  • Overview on data mining related Python librariess (spicy, numpy, pandas, scikit-learn, matplotlib). 
  • Practical case studies exploring:
    • Summarizing and visualizing multidimensional data (Exploratory Data Analysis).
    • Supervised learning (e.g. classification and decision trees).
    • Unsupervised learning (e.g. dimensionality reduction and clustering).
    • The use of data mining on space syntax and GIS spatial data.
    • AI, data mining and data visualization tools in algorithm design (Grasshopper).

 

TUTOR

João Ventura Lopes, PhD candidate

João Ventura Lopes is an architect and PhD candidate (ISCTE-IUL, Lisbon). The works of his thesis explore themes focused on the dichotomy between Emergence and Composition in the urban setting, the multidimensional analysis of urban morphology, configuration and use of public open space, and the appliance of digital tools and data mining techniques that may embrace the complexity of those relations. He graduated from FAUP (Porto) in 1998, and exercises since then professional activity on architectural and urban design, both in collaboration as individually. Post graduated in 2013 from the Course of Advanced Studies in Digital Architecture (ISCTE-IUL + FAUP).

 

PREVIOUS KNOWLEDGE

No specific prerequisites are needed, but remembering some of the basic statistics from high school (like mean, median, mode or standard deviation), and have a glimpse on Python (or other programming/visual/scripting language) and GIS or space syntax can improve the experience.

 

FORMAT

Workshop with 8 hours, in one day.

 Each trainee should bring its laptop, with the Anaconda standard installation ready to run (more instruction will follow). It will be used mostly preinstalled and space syntax (Depthmap) datasets, but participants are invited to bring their own datasets.

 

LANGUAGE

English.

The tutorial monitoring of each student can also be held in Spanish and Portuguese.

 

NUMBER OF TRAINEES

The course will take place with a minimum of 6 and a maximum of 20 trainees.