Skip to content

Latest commit

 

History

History
40 lines (23 loc) · 2.28 KB

README.md

File metadata and controls

40 lines (23 loc) · 2.28 KB
Raw

Naive Bayes Classifier on Abalone Dataset

Bahcesehir University CMP5130 MACHINE LEARNING AND PATTERN RECOGNITION Homework #1 - Naive Bayes Classifier

You are asked to implement naive bayes classifier on abalone dataset.

Note: Do not use a library or use a source code from internet, implement it yourself!

Detailed information about abalone dataset can be found at http://archive.ics.uci.edu/ml/datasets/Abalone

The aim of the dataset is to predict the age of abalone from physical measurements. Originally it is a regression problem in which the output is age in years. However, we will use it as a classification problem. The age value is already discretized as young, middle-aged, and old. The dataset (input features and class labels of the samples) is provided as a seperate text file (abalone_dataset.txt).

input: Sex,Length,Diameter,Height,Whole weight,Shucked weight,Viscera weight,Shell weight

output: class label (less than 8 in age belongs to class 1 (young), between 8 and 12 to class 2 (middle-aged), greater than 12 to class 3 (old))

Optimization on validation set is not required in Naive Bayes classification.

So, the dataset will be divided into training and validation sets only (i.e. there will not be a test set).

You are asked to implement naive bayes classifier for the following four cases:

  1. Assume gaussian distribution for continuous features. Report the accuracies for each of the following case:

1.1) 100 samples for training, and rest for validation set 1.2) 1000 samples for training, and rest for validation set

  1. Use Histogram Estimator for each of the continuous feature. Determine the bin size for each feature with your own criterion. Report the accuracies for each of the following case:

2.1) 100 samples for training, and rest for validation set 2.2) 1000 samples for training, and rest for validation set

For each of the above cases,

  • Report how many total misclassification errors are there on the training and validation sets, together with the confusion matrices. (Note: A confusion matrix is a 3x3 matrix (if # of classes is 3) where entry (i,j) contains the number of instances belonging to i but are assigned to j; ideally it should be a diagonal matrix.)

  • Report the case in which highest accuracy is obtained. Write your comments about the results.