General Info

Teachers

Teaching Assistants

Lectures

  • Tuesdays 18:00-19:45.
  • Exercise sessions

  • Tuesdays 20:00-22:00.
  • Location and TA coverage

    Materials

    Most topics will be based on the book "Mining of Massive Datasets" (Third edition) - see book homepage. However, further additional materials will be used. Observe that there is a huge community utilizing algorithms and further developing the tools we are considering within this course, so online search can also bring you valuable information and guidance.

    Videocast

  • Videocasts
  • Project work

    See information (including important dates) about the project in this PDF file. You can also look at the following example reports: Example 1, Example 2, Example 3, Example 4, Example 5.

    NOTE: You must complete the project in groups of size 4 or 5. You must form these groups on DTU Learn by October 23 at 23:59. Any groups that have fewer than 4 members after this point will be merged with other undersized groups or individuals to make a group of size 4 or 5. You will not have a choice about which groups are merged. If necessary, some undersized groups may even have to be split up in order to make all of the groups of an appropriate size.

    Weekplan

    THIS SCHEDULE IS TENTATIVE AND SUBJECT TO CHANGE

    Week Topics Slides Exercises Materials
    W1 / Sep 01
  • Introductory lecture
  • What is Data Mining?
  • Bonferroni's Principle
  • Tf.idf measure
  • Hash functions
  • Slides

  • Exercises: 1.2.1, 1.2.2, 1.3.1, 1.3.2 and 1.3.3
  • Ch. 1 of MMDS;
    W2 / Sep 08
  • No lecture
  • Python recap
  • Setting up Python and Jupyter Notebook on your local environment
  • Working on tutorial tasks
  • Tutorial tasks for NumPy, SciPy and Numba packages
  • Tutorial tasks for Pandas package
  • Exercise sheet
    W3 / Sep 15
  • MapReduce
  • Distributed File Systems
  • Cluster Computing
  • Slides Exercise sheet
    Solutions
    Ch. 2 of MMDS
    Test Files
    W4 / Sep 22
  • Similar Items
  • Minhashing
  • Locality Sensitive Hashing
  • Slides Exercise sheet
    Solutions
    Ch. 3 of MMDS
    Data and Template
    W5 / Sep 29
  • Frequent itemsets
  • Market-Basket Model
  • Association Rules
  • A-Priori Algorithm
  • PCY Algorithm (+ refinements)
  • Slides Exercise sheet
    Solutions
    Ch. 6 of MMDS
    W6 / Oct 06
  • Clustering
  • Hierarchichal algorithms
  • Point assignment algorithms (k-means algorithm)
  • DBSCAN algorithm
  • CURE algorithm
  • Evaluating (e.g. Davies-Bouldin index)
  • Slides Exercise sheet
    Solutions
    Ch. 7 of MMDS
    Holidays Holiday week
    W7 / Oct 20
  • Mining Social-Network Graphs
  • Betweenness centrality
  • Girvan-Newman algorithm
  • Modularity
  • Louvain Algorithm
  • Spectral clustering
  • Slides Exercise sheet
    Solutions
    Ch. 10 of MMDS
    Survey on Spectral Clustering Stanford lecture notes on community structure in networks by Leskovec
    W8 / Oct 27
  • Project Work
  • W9 / Nov 03
  • Project Work
  • W10 / Nov 10
  • Project Work
  • W11 / Nov 17
  • Project Work
  • W12 / Nov 24
  • Project Work
  • W13 / Dec 01
  • Project Work
  • Exam period THERE IS NO EXAM IN THIS COURSE