Supervised Learning

Question 1

Agent is given IO pairs and learns function that maps from I to O

Answer 1

Supervised learning

Question 2

Input output pairs

Answer 2

Data

Question 3

One row of input data is a

Answer 3

Feature vector

Question 4

Outputs of each feature vector are

Answer 4

Target labels

Question 5

Data fed to the agent

Answer 5

Training data

Question 6

Process of learning

Answer 6

Training

Question 7

Given that everything else is equal, choose the less complex hypothesis

Answer 7

Occam’s / ockham’s razor

Question 8

Downside of overgeneralizing

Answer 8

Function will group parts of data that are not actually related

Question 9

Downside of overfitting

Answer 9

Function will only be correct on TRAINING DATA and not on new data

Question 10

Document representation scheme that countd frequency of words in a set of documents

Answer 10

Bag of words

Question 11

Number of unique worfs across all samples

Answer 11

Dictionary size

Question 12

Which is better? Determining by considering the EXACT EMAIL or by COMPUTING probability of words in an email?

Answer 12

The latter

Question 13

Accomodating ONLY THE WORDS in the TRAINING DATA

Answer 13

Overfitting

Question 14

Introduces k fake data to prevent overfitting

Answer 14

Laplace smoothing

Question 15

Used to find the best value for the smoothing factor by PARTITONING training data

Answer 15

Cross validation

Question 16

How many percent of the data set will be used for training the agent

Answer 16

80

Question 17

How many percent will be used to compute the value of the SMOOTHING FACTOR k

Answer 17

10

Question 18

How many percent sill be used to TEST if the parameters are correct and check the accuracy of the agent

Answer 18

10

Question 19

Examples of advanced spam filters

Answer 19

Knowing the IP Address of sender
Past exchanges of email with sender
Is the email all caps
Do the URL point to where they say
Is the body and other details consistent

Question 20

Spam filtering problem has a ___ number of target labels

Answer 20

Discrete

Question 21

Fits a curve to a certain degree to a given set of TRAINING DATA

Answer 21

Regression

Question 22

Fits a line to a given training data

Answer 22

Linear regression

Question 23

Formula for linear regression

Answer 23

f(x) = w1X + w0

Question 24

The linear regression formula aims to minimize the

Answer 24

Loss function

Question 25

Computes the RESIDUAL ERROR after fitting the linear function

Answer 25

Loss function

Question 26

Algorithm taht uses linear functons for classifications

Answer 26

Perceptron algorithm

Question 27

Earliesy model of the human neuron

Answer 27

Perceptron

Question 28

Perceptron algorithm aims to find a ___ that will divide inouts to classes

Answer 28

Separator

Question 29

Methods whose parameters are finite and independent of the training size

Answer 29

Parametric methods

Question 30

Methods whose pArameters grow sd the number of training data increases

Answer 30

Nonparametric methods

Question 31

Why is k nearest ne9ghbors algorithm tedious if data is large?

Answer 31

Complexity of the search also increases

Question 32

One of the problems that can be solved using supervised learning wherein an agent analyzes example SPAM emails and HAM emails

Answer 32

Spam filtering

Question 33

Number of occurrences of Ham messages in the data set over the total number of messages

Answer 33

P(Ham)

Question 34

Probability that the email occurs in the Spam data set

Answer 34

P(email|Spam)

Question 35

A case of over fitting in spam filtering

Answer 35

Naive bayes classification

Question 36

Formula for w0

Answer 36

w0 = 1/N(sum of y) - w1/N(sum of x)

Question 37

Formula of w1

Answer 37

w1 = N(sum of xy) - (sum of x)(sum of y) / N(sum of x^2) - (sum of x)^2