Quiz
Module 6 Review
Create a free account to finish this quiz
Save your progress and see your final score.
Already have an account? Sign in
Questions in this quiz (14)
1 Which of the following is a correct validation fit statistic when predicting Decisions?
- AAll the above
- BAverage Profit
- CKolmogorov-Smirnov Statistic
- DMisclassification
2 Among multiple neural network model, __________ (validation fit statistic) value should be ____________.
- AAverage Squared Error, smallest
- BMisclassification, largest
- CLog-Likelihood, smallest
- DGini Coefficient, smallest
3What are some of the ways to improve the Default Neural Net Model?
- AChanging the architect of the network
- BChanging the output variables
- CChanging type of Neural Net(NN) from RBF to MLP
- DAlways keep the number of hidden layers low
4Which of the following are true about RBF(Radial Basis Functions)?
- ATransfer function in RBF is typically Gaussian(bell-shaped)
- BThe nodes input is a non-linear function of how close the hidden layer is to the target.
- CRadial means all inputs that are different distance from the node’s position that produce different output.
- DUnlike MLP, there are usually weights between input and output layer. Instead, this layer has RBF node’s position, width and height
5 In RBF Neural Network the weights will be applied between?
- AHidden layer and output layer
- BInput layer and hidden layer
- CCan be applied anywhere between input layer and output layer
- DNo weights are applied
6Which one is the transfer function typically used in RBF?
- AGaussian
- BUnit step (threshold
- CSigmoid
- DPiecewise linear
7Which of the following is a correct validation fit statistic when predicting estimates?
- AAverage Squared Error
- BROC Index
- CKolmogorov-Smirnov Statistic
- DMisclassification
8Based on the ROC Index, which model would be considered a strong one?
- AGreater than 0.7
- BGreater than 1
- CLess than 0.6
- DBetween 0.6 and 0.7
9In a neural network model, what changes to the input variables improves the model performance?
- ACombining or re- coding different levels in a categorical variable
- BAdding additional variables without a variable selection step
- CUsing interval variable without transformation even if the data is skewed
- DRemoving combination of existing variables
10When additional nodes are added to a Neural Network with Radial Basis Function, what shape does the response surface takes in?
- AAny shape
- BBell Shape
- CElliptical shape
- DSpherical shape
11A weak ROC Index has a threshold off?
- ARoc Index< 0.6
- BRoc Index> 0.7
- CRoc Index< 0.7
- DRoc Index> 0.6
12RBF places the redial basis (circles) functions at what locations?
- ADifferent locations in the data space
- BSame locations in the data space
- COne location in the database
- DEmpty location in the database.
13What is perceptron?
- Aa single layer feed-forward neural network with pre-processing
- Ban auto-associative neural network
- Ca double layer auto-associative neural network
- Da neural network that contains feedback
14What is true regarding backpropagation rule?
- Aall of them
- Bit is also called generalized delta rule
- Cerror in output is propagated backwards only to determine weight updates
- Dthere is no feedback of signal at any stage
Information
- Description
- Quiz on Module 6 Review, created by Nick Feldman on 01/05/2022.
- Author
- Nick Feldman
- Language
- English
- Views
- 0
- Rating
- No ratings yet Rate this quiz