Test
Unstructured Data Analytics Quizes
Este test tiene tipos de pregunta que se abren en el reproductor de siempre.
Preguntas del test (60)
1Which of these word patterns will the regular expression "^[aeiou]{2}" match?
- A"lychee" "physalis" "strawberry"
- B"bread" "wallet" "drum"
- C"apple" "physalis" "egg"
- D"eiffel" "oath" "eight"
2What is the difference between stemming and lemmatization?
- ANothing. They mean the same thing.
- BStemming is a follow-on process to lemmatization that takes part-of-speech into consideration.
- CStemming is a subset of the lemmatization process.
- DLemmatization is an advanced form of stemming that takes context into consideration.
3Which of the following text-preprocessing steps can the unnest_tokens() function do?
- Astemming
- Bremoving stop words
- Cspelling normalization
- Dsentence boundary detection
4An anti-join between table 1 and table 2 will return ___________.
- Aall records in table 1 that don't match the records in table 2
- Ball records in both table 1 and table 2
- Call records in table 1 that match the records in table 2
- Dall records in table 1 that either match or don't match the records in table 2
5The syntax for regular expressions can be grouped into four major categories. Which of these is not one of them?
- AOperators
- BEscape sequences
- CAnchors
- DMetacharacters
6JSON stands for Javascript Object [blank_start]Notation[blank_end].
- ANotation
7Data with a self-describing structure and which doesn't follow a pre-defined form can be classified as ______________ data.
- Asemi-structured
- Bpartially structured
- Cstructured
- Dunstructured
8XML (Extensible Markup Language) is a human and machine readable markup language which is similar to and uses the same set of tags as HTML.
- AVerdadero
- BFalso
9The special characters used in regular expressions to describe patterns in text are known as _____________.
- Aubercharacters
- Bmetacharacters
- Cregex
- DXML
10Which of these word patterns will the regular expression "^w.*t$" match?
- A"tight" "want" "west"
- B"wait" "want" "west"
- C"wait" "wail" "wand"
- D"wait" "mast" "west"
11Which of these best describes the document frequency of a term?
- AThe proportion of terms that appear in a document.
- BThe ratio of documents in which a term occurs.
- CThe number of times that a term appears in a document.
- DThe frequency of terms in a document corpus.
12The bag-of-words assumption for the vector-space model stipulates that words should be listed in the same order in which they appear in the original document.
- AVerdadero
- BFalso
13The process of breaking up text into individual lexical units is known as ____________.
- Aunnesting
- Btokenization
- Clexical generation
- Dn-gram generation
14In the vector space model, rows represent documents and columns represent ____________.
- Atokens
- Bspace
- Clexicons
- Dcounts
15You are presented with a regex string pattern as follows "^\\\\". Which of these best describes the pattern?
- AWords that begin with a double backslash ("\\").
- BWords that have a double backslash ("\\").
- CWords that begin with a single backslash ("\").
- DWords that end with a single backslash ("\").
16The text analytics approach that is focused on the construction of structured data, such as names, places and organizations, from unstructured data is known as _________________.
- Asearch and information retrieval
- Binformation extraction
- Cconcept extraction
- Dsyntactic parsing
17A token can either be an n-gram, sentence or paragraph.
- AVerdadero
- BFalso
18An n-gram is a contiguous sequence of n words within text. Which of these is another name for an n-gram with n set to 1?
- Aa single gram
- Ba solo gram
- Ca word
- Da phrase
19The text analytics approach that is focused on assigning labels to unlabeled documents using a model learned from documents with known labels is called ___________________.
- Adocument segmentation
- Bdocument clustering
- Csentiment analysis
- Ddocument classification
20Words that occur a lot within a document, and also occur in a lot of documents within the corpus, will have high tf-idf.
- AVerdadero
- BFalso
21The LDA() function in the R topicmodels package supports two topic modeling methods. One is "Gibbs Sampling." What is the other called?
- AVEM
- BpLSA
- CTextRank
- DLSI
22The argument syntax for the LDA() function in the topicmodels R package is is as follows: LDA(dtm, k, method, control). What does the "k" stand for?
- AThe number of topics.
- BThe number of documents.
- CThe random number seed.
- DThe alpha input vector.
23In sentence-level sentiment analysis, intensifiers, downtoners, adversative conjunctions, and negators are collectively known as valence shifters.
- AVerdadero
- BFalso
24Sentiment analysis is not impacted by the cultural and demographic differences between authors.
- AVerdadero
- BFalso
25Sentiment analysis is not only concerned with opinions about an entity, it is also concerned with opinions about the attributes of the entity. What are these attributes called?
- Aprospects
- Baspects
- Ctargets
- Dholders
26In topic modeling, there are several approaches to finding the right number of topics. Which of these is not one of them?
- AGibbs Sampling
- BTopic Coherence
- CLog-likelihood
- DPerplexity
27Given the sentence, "I don't like taking the SouthShore, but I love coming to Chicago." What type of valence shifter is the word "but"?
- ADe-amplifier
- BAmplifier
- CNegator
- DAdversative Conjunction
28Which of these is not one of the sentiment lexicons available through the "tidytext" package?
- ALoughran
- BBing
- CAFINN
- DPlutchik
29Which of these is not an assumption that the LDA model makes?
- AEvery topic is a mixture of words.
- BDocuments with similar topics will use similar groups of words.
- CEvery document is a mixture of topics.
- DEvery word is a mixture of documents.
30Dirichlet distributions are sampled over a probability simplex. A probability simplex is a set of numbers that add up to 1.
- AVerdadero
- BFalso
31A collection of words that occur together is called a
- Atopic
- Bdocument
- Csentiment
- Dword
32Which of these is not a common technique used for topic modeling?
- ALatent Dirichlet Allocation
- BProbabilistic Latent Semantic Analysis
- CLatent Semantic Indexing
- DLatent Syntactic Parsing
33Given the sentence: "I don't like taking the SouthShore but I love coming to Chicago." What type of valence shifter is the word "don't"?
- AAdversative Conjunction
- BAmplifier
- CNegator
- DDe-amplifier
34Topic models are mixture models. This implies that a particular document can be assigned to more than one topic. In clustering, what kind of clustering approach would this be called?
- ASoft clustering
- BMixture clustering
- CExclusive clustering
- DComplete clustering
35The tidy text word-level sentiment analysis approach takes into consideration qualifiers, such as "pretty good", "not bad", etc.
- AVerdadero
- BFalso
36The valence of an emotion, that is, the degree to which it is positive or negative, is also known as the _________ of the emotion.
- Alexicon
- Bamplifier
- Cpolarity
- Dframework
37The LDA() function from the topicmodels package returns two matrices. One is called gamma and the other is called beta. What does beta represent?
- AThe prevalence of topics in documents.
- BThe probability of words belonging to documents.
- CThe probability of words belonging to topics.
- DThe prevalence of words in documents.
38All four of the tidy text sentiment lexicons are based on unigrams.
- AVerdadero
- BFalso
39Unlike the tidytext package, the sentimentr package does not allow the use of custom sentiment lexicons.
- AVerdadero
- BFalso
40Dirichlet distributions are parameterized by a vector (alpha). As alpha __________ the distribution becomes more spread out and moves towards the center. As alpha ____________ the distribution moves towards the poles.
- Adecreases, increases
- Bbecomes positive, becomes negative
- Cincreases, decreases
- Dbecomes negative, becomes positive
41Because events cannot simultaneously happen and not happen, events are mutually exclusive with their complements.
- AVerdadero
- BFalso
42Given two independent events A and B, with P(A) = 0.5 and P(B) = 0.4. The joint probability P(A ⋂ B) =
[blank_start]0.2[blank_end].- A0.2
43For dependent events A and B, the probability of event A, given that event B occurred is described with the conditional probability P(A | B) = P(A ⋂ B) / P(B). This is known as the ______________.
- AConditional Probability Theorem
- BBayes Theorem
- CBarnes Theorem
- DTesla Theorem
44In the conditional probability formula above, P(spam | free) is known as the _________.
- Aposterior probability
- Bmarginal likelihood
- Clikelihood
- Dprior probability
45The Naive Bayes algorithm is premised on the “naive” assumption that all features in the dataset are ________ important and __________.
- Adependably, equal
- Bunequal, undependable
- Cequally, dependent
- Dequally, independent
46"Events are independent so long as they are conditioned on the same class value." This is known as __________________.
- Avalue independence
- Bclass-conditional independence
- Cbayesian independence
- Dcross-conditional independence
47With Naive Bayes, the method of adding a pseudo-count to the number of observed and unobserved cases in order to change the expected probability so as to avoid zero-frequency problems is known as _____________.
- APseudo-smoothing
- BLaplace smoothing
- CBayesian smoothing
- DNaive smoothing
48One of the weaknesses of the Naive Bayes algorithm is that it assumes that features are equally important and independent.
- AVerdadero
- BFalso
49Similar to logistic regression, Naive Bayes models are ____________ models.
- Aparametric
- Bmetric
- Cmatrix
- Dnon-parametric
50Which of these is not a strength of the Naive Bayes algorithm?
- AIt is not affected by correlated features.
- BIt is useful for both continuous and discrete features.
- CIt works well with both little or large training data.
- DIt works well with noisy and missing data.
51The hyperplane that creates the largest separation between classes is known as the ________________ hyperplane.
- Asupport vector
- Bkernel
- Cmaximum margin
- Dbest fit
52In the support vector machine approach, each class must have at least one support vector.
- AVerdadero
- BFalso
53Which of these is not an approach used in SVMs when dealing with non-linearly separable data?
- AUsing a technique known as soft-margin classification.
- BIntroducing a slack variable and allowing some examples to be misclassified.
- CUsing a kernel transformation.
- DNormalizing the data so it fits within the margin of the hyperplane.
54Which of these is not a commonly used kernel transformation?
- AHyperplane Kernel
- BSigmoid Kernel
- CGaussian Kernel
- DLinear Kernel
55Which of these is not a weakness of support vector machines?
- AThey can only be used for classification problems.
- BThe choice of kernel has a significant impact on the performance of the model.
- CIt's a black box approach.
- DThe results typically do not have a direct probabilistic interpretation.
56Support vector machines are not good against high-dimensional datasets.
- AVerdadero
- BFalso
57With support vector machines, the process of transforming data to a higher dimensional space in order to make it linearly separable is known as ______________.
- AThe kernel trick.
- BHigh-dimensional classification.
- CThe support vector transformation.
- DThe maximum margin hyperplane.
58Which of these is not a strength of support vector machines?
- AThey are not overly influenced by noisy data.
- BThey are useful for both classification and regression.
- CBecause of the kernel transformation, they are very fast to train.
- DThey are not prone to overfitting.
59When classifiying data with support vector machines, only a subset of the data is used in the process. Why is this?
- AThe kernel transformation gets rid of most of the data.
- BOnly the support vectors are used in the classification process.
- CThe kernel trick reduces the dimensionality of the data.
- DThe sampling approach is highly reductive.
60Both Naive Bayes and Support Vector Machines are unsupervised machine learning algorithms.
- AVerdadero
- BFalso
Información
- Descripción
- Final Machine Learning Test sobre Unstructured Data Analytics Quizes, creado por Kelly Cronin el 04/10/2019.
- Autor
- Kelly Cronin
- Idioma
- English
- Visitas
- 176
- Etiquetas
- Valoración
- Aún sin valoraciones Valora este test