Test
Module 5 Review
Crea una cuenta gratis para terminar este test
Guarda tu progreso y consulta tu resultado final.
¿Ya tienes cuenta? Inicia sesión
Preguntas del test (16)
1What is typically the best type of model?
- ASimplest Model
- BComplex Model
- CIntricate Model
- DMachine Learning Model
2Which term describes how much an input must change to double the odds for the desired output?
- ADoubling Amount
- BOdds Ratio
- CPoint Estimate
- DLeverage Point
3Which of the following methods do we use to best fit the data in Logistic Regression?
- AMaximum Likelihood
- BJaccard distance
- CLeast Square Error
- DBoth A and B
4One of the very good methods to analyze the performance of Logistic Regression is AIC, which is like R-Squared in Linear Regression. Which of the following is true about AIC?
- AWe prefer a model with minimum AIC value
- BWe prefer a model with maximum AIC value
- CBoth depend on the situation
- DNone of these
5What is the best model selection criteria when the goal is classification(decisions) in Logistic Regression
- ALow Misclassification Rate value
- BValidation error
- CAIC
- DNone of these
6Why do we need coding consolidation for non-numeric input variables
- ATo overcome the problem of exploding dimensions
- BTo make data size less
- CTo increase model complexity
- DNone of these
7How do we handle high leverage points in an input (X) variable for LR model?
- ATransformation of input variables
- BExclude high leverage points from LR model
- CBoth
- DNone of these
8The logit link function transform probabilities ( between 0 to 1) to logit score between ………………………….
- A- ∞ to ∞
- B-1 to 1
- C0 to ∞
- D-1 to 0
9Which of the following steps help in selecting useful inputs for regressionn
- ASequential Selection
- BPrediction Formula
- CBest Model Selection from sequence
- DNone of these
10Which of the following functions signify secondary outcome training cases
- AΣlog(1−pi)
- BΣlog(pi)
- CΣlog(1+pi)
- DΣlog(1/pi)
11Regarding splits in Decision Trees what criteria makes a great split?
- AThe best split is one that increases purity of the sub-sets by the greatest amount
- BA good split also creates nodes of similar size or at least does not create very small nodes
- CThe weighted sum of items for each split has the smallest amount.
- DA good gini or enthropy split
12What is a Common issues in Deep Learning and Machine learning?
- ARequires clean data
- BBe careful not to over fit or, under fit data during training
- CChoose hyper parameters carefully
- DChoice of cost function
13The cost function used in linear regression work in logistic regression.
- AFalse
- BTrue
14Using -------------- Values we build the ROC (Receiver Operator characteristic) curve.
- AConfusion matrix
- BAIC value
- CMultiple AIC Values
- DNone of these
15How much does an ‘input’ need to change to double the odds?
- A0.69/wi = odds*2
- B0.59/wi = odds*2
- C0.49/wi = odds*2
- D0.39/wi = odds*2
16How do we calculate odds ratio?
- ALog(p̂/(1- p̂))
- BLog((1- p̂)/ p̂)
- CLog(p̂/(1* p̂))
- DLog((1* p̂)/ p̂)
Información
- Descripción
- Test sobre Module 5 Review, creado por Nick Feldman el 01/05/2022.
- Autor
- Nick Feldman
- Idioma
- English
- Visitas
- 2
- Valoración
- Aún sin valoraciones Valora este test