22) The determinant features of the 3-class X dataset to which the C4.5 decision tree learning algorithm will be applied

[answerhappygod]

22)
The determinant features of the 3-class X dataset to which the C4.5
decision tree learning algorithm will be applied and the
categorical values they take are given below. Which rule cannot be
learned with this algorithm?
O1 = {a,b,c,d}
O2= {x,y,z}
CLASS={S1,S2,S3}
a. (O1=b) OR (O2= not y) then CLASS=s2
b. All can be learned
c. If O1=d CLASS=S2
d. CLASS=S3 if O2=z
e. If O1=c AND O2=y CLASS=S3
------------------------------------------------
23)
Which of the following is not a desirable feature of any clustering
algorithm?
a. Data is not dependent on the order of arrival.
b. Compatible with different data types
C. Requires domain knowledge
d. Being declarable in terms of time
e. Scalable in terms of memory
------------------------------------------------------------
24)
A and B fuzzy sets consist of exactly the same elements. The
membership degrees of the elements to these sets are as
follows:
A={0.3, 0.7, 0.5, 0.2}
B={0.6, 0.9, 0.45, 0.2}
Accordingly, in which option are the membership degrees of the
elements forming the AUB fuzzy set correctly given?
a. {0.9, 1.6, 0.95, 0.4}
b. {0.3, 0.7, 0.45,0.2}
c. Incalculable
d. {0.6, 0.9, 0.5, 0.4}
e. {0.6, 0.9, 0.45, 0.2}
----------------------------------------------
25)
Which statement about noisy and outlier data is true?
a.Noisy data is the same as counterfeit data
b. Outlier data is only data that has very, very small values
compared to its peers.
c. Outlier data has been measured correctly
d. All
e. Noisy data cannot be caused by sensor errors
------------------------------------------
26)
Which of the following is not a desirable feature of any clustering
algorithm?
a. Compatible with different data types
b. Requires domain knowledge
c. Scalable over time
D. Data is not dependent on the order of arrival.
e. Memory scalable
27)Which of the following statements is true?
a. A person's degree of membership in the concept X and the
probability of being a member of the concept in question are
different concepts.
b. All
c. If a person's degree of membership in the concept X in the
society is 0.4, that person does not have the concept X with a 60%
probability
D. If the probability of a person in the society to have concept
X is 10%, that person's degree of membership to concept X is
0.1.
e. If a person's degree of membership in the concept X is 0.5,
that person has a 50% probability of having the concept X.
28) What is the distance, in units, between the words "fener"
and "yazar" in an environment where substitution, insertion,
deletion operators are allowed?
a. 3
b. 10
c. 4
D. 5
e. 9
29)
In what range does the membership function of a concept take values
according to fuzzy logic principles?
a)[0,1)
b)[0,1]
c)(0,1]
d)(0,1)
e){0,1}
---------------------------------------------
30) Under what condition is the minimum entropy achieved in a
population with a maximum of 2 classes?
a. If the population consists of 25% of one class and 75% of the
other class
b. None
C. If the population consists of only one class
d. If the population is evenly distributed between 2 classes
e. If the noisy data rate in the population is 50%
------------------------------------------------------
31) In which option is the element(s) of the generalization
error made by a machine learning model correctly given?
a. Just Bias
b. Bias, Variance
c. Variance, Standard Deviation
D. Bias, Standard Deviation
e. Variance Only
===========================
32)The defining features and categorical values ​​of the 2-class
X dataset to which the decision tree learning algorithm will be
applied are given below. Which rule cannot be learned with this
algorithm?
S1 = {a,b,c}
S2 = {x,y}
CLASS={s1,s2}
a. None
b. If S1=x then CLASS=s1
c. If S1=y AND 02=b CLASS=s2
D. If S2=a CLASS=s2
e. If S1=x OR S2=a CLASS=s1
==============================
33)Which statement about training and test error is true?
a. Regardless of the complexity of the model, the test error is
higher than the training error
b. The training error never converges to zero.
c. None
D. Test error can exceed 100%
e. The complexity of the model has nothing to do with training
and test error.
==============================
34)What is the maximum entropy of a dataset with 4 classes?
a. 0
b. 2
c. 4
D. None
e.1
=============================
35) A machine learning algorithm is trained with 100 different
training data and 100 different models are obtained. The average of
these 100 different models is then taken. Which option best
expresses the reason why the average model is different from the
real model we are trying to reach?
a. None
b. The mean model contains false assumptions and
simplifications.
c. The mean model contains simplifications
D. The mean model contains false assumptions
e. Insufficient testing of the mean model
===============================
36) If late pruning is used in the decision tree algorithm, how
many parts should the original data be divided into at first?
a. None
b. 2
c. It doesn't matter
d. 3
e. 1
============================
37) There are 90 records in a training data consisting of 3
classes. Which of the following values ​​of K should not be chosen
according to the K-near neighbor algorithm?
a. 7
b. 9
c. 8
D. 5
e. 4
--------------------
38) What is the distribution of errors on training and test data
in a simply designed model?
a. Varies depending on data
b. Low training error, low test error
c. High training error, high test error
D. Low training error, high test error
e. High training error, low test error
--------------------
39) As the number of nodes added to a decision tree increases,
the accuracy on the test data first decreases and then increases
again.
Select one of them:
1) Correct
2) Wrong
------------------------
40) In the model designed with too few and too many parameters,
what is the distribution of bias, respectively?
a. Small big
b. big, big
c. Big small
D. Varies according to the content of the data
e. Little little
-----------------
41) How many association rules are created from a frequent
product set of 5 elements in an environment where minimum accuracy
criteria are not sought?
a. 32
b. 4
c. 30
D. 10
e. 5
-------------------
42) In the decision tree learning algorithm, how many times can
the X determinant attribute of the data set be included in the
tree, minimum and maximum? (Let the defining attributes of the
dataset be X, Y, Z, V, W)
a. Minimum:0 Maximum:2
b. Minimum:1 Maximum:1
c. Minimum: 1 Maximum:5
D. Minimum:0 Maximum:1
e. Minimum:0 Maximum:5
-----------------------
43) Which of the following occurs as the complexity of a model
increases?
a. Test error tends to increase first, then decrease
b. The training error tends to increase continuously.
c. Going from high variance to low variance
D. Going from high bias to low bias
e. None
----------------------
44) When classifying according to the Naive Bayes classifier,
the fact that some product terms are zero may upset all
probabilistic calculations. What is the main reason for this?
a. Homogeneous distribution of training data among classes
b. Not using Naive Bayes to find relationship rules
c Insufficient number of personnel of the training data
D. Naive Bayes being a weak classifier
e. None