In this problem the Iris Plants Database contains 3 classes of 50 instances each, where each class refers to a type of Iris plant. The attributes/features were extended by two features for each plant instance for a total of 6 features. The updated data is provide as a csv file. In this assignment data processing and machine learning techniques need to be implemented as follows:
- Data Cleansing (use the iris data for cleansing.csv)
- Data Transformation (can be combined with #6 (dimensionality reduction))
- Generate two sets of features from the original 4 features to end up with a total of 8 featres
- Perform Feature Preprocessing
Use an outlier removal method to remove any outliers.
- Rank the 6 set of features to determine which are the two top features
- Reduce the dimensionality to two features using PCA or Kernel PCA
- Using the following Machine Learning techniques, classify the three class Iris data:
- Expectation Maximization
- Either Fisher Linear Discriminant (Linear Discriminant Analysis), Kernel Fisher’s Discriminant or Parzen Window
- Neural Network Method (Probabilistic NN, Radial Basis Function or Feed Forward) (d) Support Vector Machine
