CertIIMKAIML

2. the developed model is such that the user can only test the model but cannot know how it is working
3. In this case the training data does not include any desired output
4. error related to misestimating the mean
6. the process that is used to classify nodes into two or more sub-nodes
11. Selecting right model
12. used to control the learning process
13. A method that use the inbuilt structures in the data for a better organization of the data
14. Computing linear correlation after ranking the variable values

1. Using many models to generate the estimate of the required categories Normally used any time you use a common noun
5. Transforms all the values of any attribute to zero mean and standard deviation one
7. Correcting/removing inconsistencies, inaccuracy, irregularities etc in the data
8. this problem can be solved by using constraints on model parameters
9. fitting one example to a known class
10. learns through the induction process