Font size
WorksheetsML
Total questions: 48
Worksheet time: 48mins
_________is programming computers to optimize a performance criterion using example data or past experience.
ML
Data science
Big data
Neural network
Which of these doesn't require 'learning'
navigating on Mars
speech recognition
routing on a computer network
user biometrics
calculate payroll
Which of these are the applications of classification in ML
Pattern recognition
Face recognition
Outlier/novelty detection
Medical diagnosis
Character recognition
Examples of unsupervised learning
Customer segmentation in CRM
Image compression: Color quantization
Bioinformatics: Learning motifs
Outlier detection: Credit card Fraud
(a) predicts the value of the dependent attribute from the attribute variables
Dependent attribute is (a) for regression
Dependent attribute is _______for classification
Numerical
Alphanumeric
Categorical
Mixed data
Different types of data in a dataset
Transactional Data and Sequence Data
Spatio-Temporal Data
Interval Data and Time Series Data
Spatial Data and Data set with multiple kinds of data
Interpretations of noise
Imprecision in recording the input attributes
Errors in labeling the data points
Additional attributes which were not considered
Unwanted anomaly in the data
Abbreviate PAC
Probably Approximately Correct
Principle Component Analysis
Principle Analysis of Components
______ is the process of choosing one among many candidate models for a predictive modeling problem.
Model Extraction
Model Creation
Model Generalization
Model Selection
There are ______ways to write 'd' binary values in the Model selection and generalization with a binary function
2^2d
2^d
d^n
n^d
______set of assumptions we make to have learning possible
Variance
Bias
Inductive bias
Hypothesis class H is less complex than the function
Good balanced model
Underfitting
Noise
Overfitting
________tries to find the parameter values that maximizes the likelihood function
MLE (Maximum Likelihood Estimation)
Probability Distribution Function
Likelihood Function
Bernoulli Density
Gaussian Distribution / Normal Distribution represent data as a____
Polygon
Bell curve
Straight line
Clusters
_______is a integral of pdf over the range a & b
Probability
Profitability
Counting
Examples of Parametric methods
Regression
LDA
KNN
Simple perceptron
SVM
Disadvantages of Parametric methods
Constrained
Limited Complexity
Poor fit
Overfitting
Examples of Non parametric methods
KNN
Decision Trees
SVM
Naive Bayes
Neural Networks
Disadvantages of non parametric methods
Lots of data
Overfitting
High performance
Lots of features
_______is single global model and ________ is small number of local models
Parametric, Non parametric
Non Parametric, Parametric
Semi Parametric, Parametric
Parametric, Semi Parametric
Case-based or Instance-based distance measures
KNN
Learning Vector Quantization(LVQ)
SVM
Decision Tree
_____represents the shortest distance between two points
Minkowski Distance
Manhattan Distance
Euclidean Distance
Hamming Distance
Find the Euclidean distance between points P(3, 2) and Q(4, 1).
PQ = √2 unit.
PQ = 2 unit.
PQ = √3 unit.
PQ = √4 unit.
Find manhattan distance of
row1 = [10, 20, 15, 10, 5]
row2 = [12, 24, 18, 8, 7]
12
14
13
15
Hamming distance of
String 1 = 1V1A0R6U0NS9
String 2 = 1A1B0H6ILASH
7
7+1
5
6
This formula is used to evaluate_____
Hamming Distance
Minkowski Distance
Manhattan Distance
Bernoulli Density
____model parameters actually grow with the training data set
KNN
SVM
NAIVE BAYES
RANDOM FOREST
Chooses subset of the original set
Feature selection
Feature Extraction
Feature Creation
LDA
PCA is ____________
Dimensionality Reduction
Standardizing the data
Computing the Covariance
Multiclass Classifier
The number of instances that were positive (+) and incorrectly classified as negative (-)
Type 2 Error
Type 1 Error.
False Negative
True Positive
A measure of the variance that a factor explains for observed variables.
Eigen vector
Eigen value
Principle component
Factor
in _______ data is stretched, rotated and translated
Factor Analysis
PCA
LDA
PAC
_____centers the data at the origin and rotates the axes
PCA
LDA
Factor Analysis
PAC
Difference between truth value of parameters and mean value of the estimator is ____
Bias of the parameter
Bias of the estimator
Variance of the parameter
Bias of the parameter
Model with low bais and low variance is ___
Good model
Balanced model
Overfitting
Underfitting
Distance from the discriminant to the closest instances on either side
Boundary
Margin
Hyper plane
Support vector
SVM algorithm creates the best line or ________
Decision boundary
Hyperplane
Support vector
Margin
This best decision boundary is called a ______.
Margin
Support vector
Hyperplane
SVM
_________is a method of using linear classifier to solve a non-linear problem
Kernel trick
Kernel function
Kernel bias
Kernel machines
__________activation function takes a value as input and outputs another value between 0 and 1
Recurrent
Tanh
relu
Sigmoid
In BPN minimum value of the error function in weight space is done by _____
PCA
Factor Analysis
Gradient descent.
LDA
Calculate h1
0.2775
1.2775
1.3755
0.3775
Which of these is not an element of Reinforcement learning?
Environment
Policy
Reward
Action
Sarsa
Global Approach - Learner fusion
Voting and Stacking
Voting and Gating model
Cascading
Voting and Cascading
Generating a sequence of base learners each focusing on the previous one's errors
AdaBoost
Bagging
Voting
Cascading
Multistage Combination methods
Uses serial approach
Cascading
Stacking
Voting
Boosting
