Week 8: Data Stream Mining

In this practical, we use the stream R package for analysing stream data. Please install the stream package to complete the practical.

I. Creating a data stream

1. We firstly create a generator to generate stream data points that will belong to one of three clusters (k=3). Each data point will have 2 dimensions (d=2). The data points will follow Gaussian distribution with 5% noise. When a new data point is requested from this data generator, a cluster will be chosen randomly using the probability weights in p.

   library("stream")
   stream <- DSD_Gaussians(k = 3, d = 2, noise = .05, p = c(.5, .3, .1))
   stream

2. Generate 5 data points using the generator.

Week 11: Geospatial data analysis

Spatial objects are typically represented using vector data, which includes information about the shape or geometry of the objects, along with additional variables. For instance, a vector dataset may describe the boundaries of countries (geometry) and also store attributes like country names and population size. Similarly, it can include the geometry of roads in an area, along with attributes such as road type and names.

Week 2: Naive Bayes (Practical 1)

Part1 : Training

Steps of using Weka

Step1: Explorer

Step2: Select data

Step3: Choose Algorithm

Step3: Start and Result

Week 7: Latent Dirichlet Allocation (Practical 6)

In this practical, we will process the text data that includes the abstracts of all papers in the Journal of Statistical Software (JSS), up to 08/05/2010.
The JSS data is available as a list matrix in the package corpus.JSS.papers which can be
installed and loaded by:

Background

There are three methods to establish a classifier

A. Model a classification rule directly
B. Model the probability of class memberships given input data
C. Make a probabilistic model of data within each class