1. Apprendre
    2. /
  2. Cours
    3. /
  3. Predictive Analytics using Networked Data in R
  • 1

    Introduction, networks and labelled networks

    Gratuit

    In this chapter you will be introduced to labelled networks, network learning and the challanges that can arise.

  • 2

    Homophily

    In this chapter you will learn about homophily and how to compute the two measures that can be used to characterice it, dyadicity and heterophilicty.

  • 3

    Network Featurization

    In this chapter you will use the igraph package to compute various network features and add them to the network.

  • 4

    Putting it all together

    In this chapter you will use the network from Chapter 3 to create a flat dataset. Using standard data mining techniques, you will build predictive models and measure their performance with AUC and top decile lift.

Initializing

Exercice

Compute heterophilicity

Similar to dyadicity, heterophilicity is the ratio between the actual number of mixed label edges and the expected number of mixed label edges. In this exercise, you will compute the network's heterophilicity.

For the expected number of mixed edges, use the formula \(n_ {NC}\cdot n_ {C}\cdot p\) where \(n_ {NC}\) is the number of non-churners, \(n_ {C}\) is the number of churners and p is the connectance.

You can use the variables NonChurnNodes, ChurnNodes, connectance, and MixedEdges.

Instructions

100 XP
  • Compute the expected number of mixed edges and assign it to the variable ExpectedHet.
  • Compute the heterophilicity by dividing the actual number of mixed edges with the expected number of mixed edges. Call this variable Het.
  • Is the network heterophilic? Inspect Het to find out.