Datasets:

inez1m

/

CA400-1

like 0

When z-score transformatio n is done on continuous data, what will be the mean and standard deviation of the transformed data?

B

Please choose the appropriate answer from the following options: A. Mean 1, standard deviation 0 B. Mean 0, standard deviation 1 C. Depends on data value changes D. Insufficient data

Identify the scaling technique which re- scales a feature value so that it has distribution with 0 median value.

C

Please choose the appropriate answer from the following options: A. Standardizatio n B. Min-Max Normalization C. Robust Scaler D. Normalizer

Which of the following cannot be used to detect an outlier in the data?

C

Please choose the appropriate answer from the following options: A. Boxplot B. scatterplot C. Normalisation D. Probability plot

Imagine, you are given a list of items in a DataFrame as below. D = [‘A’,’B’,’C’,’D’,’ E’,’AA’,’AB’] Now, you want to apply label encoding on this list for importing and transforming, using LabelEncoder. from sklearn.prepro cessing import LabelEncoder le = LabelEncoder() What will be the output of the print statement below ? print (le.fit_transfor m(D))

C

Please choose the appropriate answer from the following options: A. array([0, 2, 3, 4, 5, 6, 1]) B. array([0, 2, 3, 4, 5, 1, 6]) C. array([0, 3, 4, 5, 6, 1, 2]) D. Any of these

Which of the following python code can be used to arrange 8 plots in 2 rows 4 columns?

B

Please choose the appropriate answer from the following options: A. plt.subplots(4, 2,figsize=(10,1 0)) B. plt.subplots(2, 4,figsize=(10,1 0)) C. None of these D. plt.subplots(3, 2,figsize=(10,1 0))

Which key in the boxplot output dictionary can be referred to look at the list of extreme observations?

C

Please choose the appropriate answer from the following options: A. boxes B. whiskers C. fliers D. caps

Which kind of scaling is preferred if the data is having outliers ?

B

Please choose the appropriate answer from the following options: A. Min-Max Scaler B. Robust Scaler C. Normalizer D. Standard scaling

Identify the correct code to get the percentage of missing values in the data. Assume pandas is imported as pd.

D

Please choose the appropriate answer from the following options: A. pd.isnull(data). sum()*100 B. pd.isnull(data). sum() C. pd.isnull(data). perct() D. pd.isnull(data). sum()/data.sh ape[0]*100

Which of the following statements is true for normally distributed bell curve?

B

Please choose the appropriate answer from the following options: A. Mean >= Median B. Mean=Median =Mode C. Standard Deviation = +/- 1 D. Variance = cube root of standard deviation

Aften standard scaling a data set, which of the following is not true: A) Mean of all the individual columns is approx 0 & Variance is not equal to standard deviation. B)Mean of all the individual columns is approx 0 & Variance = 1 C) Mean of all the individual columns is approx 0 and Variance is equal to standard deviation

D

Please choose the appropriate answer from the following options: A. C B. B C. B and C D. A

Which of the following is an example of univariate analysis?

D

Please choose the appropriate answer from the following options: A. boxplot B. histogram C. density plots D. All of these

What will the following command render? plt.barh(data[' region'], data['funding'] , color = 'blue') Assume Region is a categorical variable and funding a continuous numeric variable.

C

Please choose the appropriate answer from the following options: A. A horizontal bar chart with blue fill using matplotlib library for funding by region with funding on Y- axis and Region on X- Axis B. A bar chart with blue fill using matplotlib library for funding by region with funding on X- axis and Region on Y- Axis C. A horizontal bar chart with blue fill using matplotlib library for funding by region with funding on X- axis and Region on Y- Axis D. A horizontal bar chart with blue fill using pandas library for funding by region with funding on X- axis and Region on Y- Axis

Which of the following is true to calculate the upper whisker?

B

Please choose the appropriate answer from the following options: A. Q3 - 1.5*IQR B. Q3 + 1.5*IQR C. Q1 + 1.5*IQR D. Q1 - 1.5*IQR

In which of the following chart the data points are divided into portions?

C

Please choose the appropriate answer from the following options: A. Bar chart B. Column chart C. Pie chart D. Scatter chart

Consider a dataset of size 2000 x 12, where 2000 is the number of observations and 12 is the number of features. 1298 values are missing from the fourth feature column. Which would be the good strategy to handle this missing values ?

D

Please choose the appropriate answer from the following options: A. Drop the 1298 observation rows where the fourth feature column values are missing B. Impute with Mean C. Impute with Median D. Drop the whole fourth feature column

Lets say we have a Qualitative Categorical attribute in the data as “Remote” which consists of two labels as someone who works remotely or someone who does not work remotely.Whic h of the Python syntax will NOT be able to create dummy (one hot encoded) features from Categorical variables? Assume that all the standard imports are done

C

Please choose the appropriate answer from the following options: A. data["Remote _dummy "]=pd.get_du mmies(data[" Remote "]) B. data["Remote _dummy "] = data["Remote "].map(lambd a x: 1 if x == "Remote " else 0) C. data["Remote _dummy "] = data["Remote "].filter(lambd a x: 1 if x == "Remote " else 0) D. data["Remote _dummy "] = data["Remote "].apply(lambd a x: 1 if x == "Remote " else 0)

Which of the following is not true about Pearson’s Correlation Coefficient

C

Please choose the appropriate answer from the following options: A. Correlation is calculated b/w two numerical variables B. Absolute value of correlation will always lies between 0 and 1 C. Correlation can explain the non-linear relationship between two metrics D. If correlation is approximately equal to 0.9, we can expect that increase in one numerical variable will result in increase of other numerical variable also

PCA is used for

A

Please choose the appropriate answer from the following options: A. Dimensionality reduction B. Filling NA values C. Outlier Detection D. Correlation analysis

Identify the methods to handle missing values in a dataset

A

Please choose the appropriate answer from the following options: A. All of these B. Drop the missing values C. Impute with Mean/Median /Mode D. Impute with ffill /bfill

Which one of the following is not a Univariate measure ?

C

Please choose the appropriate answer from the following options: A. 75 percentile value B. Mean C. Correlation D. Standard Deviation

In box plot what does the middle line represent

C

Please choose the appropriate answer from the following options: A. First quartile B. Fourth quartile C. Second quartile D. Third quartile

Min-maxScaler is calculated by the formula

A

Please choose the appropriate answer from the following options: A. (x- xmin)/(xmax- xmin) B. (x- xmean)/(xmax- xmin) C. (x- xmax)/(xmax- xmin) D. (xmax- xmin)/(x- xmax)

The data set below has an outlier of 42. 2, 5, 12, 15, 19, 4, 6, 11, 16, 18, 12, 12, 42 What effect does removing the outlier have on the distribution of the data?

A

Please choose the appropriate answer from the following options: A. The mean will decrease B. The median will decrease C. The mean will increase D. The median will increase

Which function is used to convert from one datatype to another?

C

Please choose the appropriate answer from the following options: A. dtype B. as C. astype D. type

Identify the correct code to make horizontal bar chart using matplotlib library. Assume that the library is imported and aliased as plt. x and y are features in the dataframe.

B

Please choose the appropriate answer from the following options: A. plt.bar(x,y) B. plt.barh(x,y) C. plt.bar(x,y,vert =False) D. plt.barplot(x,y)

Which is the algorithm that works best in 90% of the cases, even when we are ignorant of the domain knowledge?

B

Please choose the appropriate answer from the following options: A. Decision Tree B. Random Forest C. Semi- supervised learning D. Linear Regression

Email filtering in gmail into Genuine (Inbox) and junk mails(promoti ons) and other categories is a type of

C

Please choose the appropriate answer from the following options: A. Regression B. Clustering C. Classification D. Apriori

The maximum Possible value of R2 (R Squared) is

B

Please choose the appropriate answer from the following options: A. 0% B. 100% C. It can have any Value - depends on the Clients preposition D. 200%

Calculate overall gini impurity after splitting usnig the following information. Samples before split=300. Left side samples after splitting=200 (150 positives, 50 negatives); Right side samples after splitting=100(5 0 positives; 50 negatives)

C

Please choose the appropriate answer from the following options: A. 0.21 B. 0.31 C. 0.41 D. 0.11

Internal nodes/sub- nodes of a decision tree correspond to:

B

Please choose the appropriate answer from the following options: A. Classes B. Attributes C. Data instances D. None of these

Bagging is built in which Algorithm?

Random Forest

Please choose the appropriate answer from the following options: A. Linear Regression B. Logistic Regression C. K-Means Clustering D. RandomForest

The x-axis and y-axis in ROC curve respectively represents:

A

Please choose the appropriate answer from the following options: A. FPR and TPR B. TPR and FPR C. FPR and TNR D. FNR and TPR

Naïve Bayes is a type of

A

Please choose the appropriate answer from the following options: A. Classification B. Clustering C. Regression D. Network Analysis

Consider the two statements stated below. Which of the following is true while deciding best split in the root node : A) Information gain should be maximum B) Gini impurity should be as minimum as possible

B

Please choose the appropriate answer from the following options: A. Both A and B are wrong B. Both A and B are correct C. A is correct; B is wrong D. A is wrong; B is correct

I am given a dataset with 20000 rows of data. Most of the data including the dependent variable is categorical and the data is not sparse. Which Classification would I chose?

A

Please choose the appropriate answer from the following options: A. Naïve Bayes B. KNN C. K-Means Clustering D. K-Median Clustering

The minimum and maximum values of the entropy are

A

Please choose the appropriate answer from the following options: A. (0 -1) B. (0 -0.5) C. None of these D. (0 .5- 1 )

KNN is this type of algorithm

C

Please choose the appropriate answer from the following options: A. Bad Algorithm B. Lethargic Algorithm C. Lazy Algorithm D. Lavish Algorithm

What does a multiple linear regression analysis examine?

A

Please choose the appropriate answer from the following options: A. The relationship between one dependent and more than one independent variables B. The relationship between more than one independent variables C. The relationship between one or more than one dependent and only one independent variable D. The relationship between more than one dependent and only one independent variable

For HR Attrition data, while predicting Attrition column using Age, MonthlyIncom e and DistanceFrom Home columns, identify gini impurity for the right node (set random_state =100, wherever necessary)

D

Please choose the appropriate answer from the following options: A. 0.6 B. 0.301 C. 0.151 D. 0.199

In a Classification Tree , the splitting of node is based on

A

Please choose the appropriate answer from the following options: A. Impurity measures such as Gini index or entropy for classification tree B. Mean Squared Error (MSE) C. Sum of Squared Errors (SSE) D. F-test

Logistic regression is used for

A

Please choose the appropriate answer from the following options: A. Classification B. Predicting continous variables C. Predicting continous and discrete variables D. Deep learning

Which of the following is a measure of accuracy

A

Please choose the appropriate answer from the following options: A. (TP+TN)/(TP+T N+FP+FN) B. (TP)/(TP+TN) C. (TP+TN)/(FP+F N) D. TP/TN

Which type of learning is most commonly used to create models on labelled categorical data(target value is also labelled)?

B

Please choose the appropriate answer from the following options: A. Regression B. Classification C. Clustering D. Componentiza tion

Which of the following is part of Data Imputation?

A

Please choose the appropriate answer from the following options: A. Filling Missing Values B. Splitting Data Set C. Adding Null Values D. Plotting Graphs

Which of the following can be called as False Negative?

A

Please choose the appropriate answer from the following options: A. Actual value is 1 and classifier predicted it as 0 B. Actual value is 0 and classifier predicted it as 0 C. Actual value is 1 and classifier predicted it as 1 D. Actual value is 0 and classifier predicted it as 1

Which is considered the best Data Validation technique?

A

Please choose the appropriate answer from the following options: A. Cross Validation B. Data Splitting C. No need of any validation - Use the model directly D. Any of the above

For HR Attrition data, while predicting Attrition column using Age, MonthlyIncom e and DistanceFrom Home columns, how many samples move to left side of the parent node (set random_state =100, wherever necessary)

D

Please choose the appropriate answer from the following options: A. 300 B. 100 C. 321 D. 233

Which is the error code prominently used in evaluating Regression Models?

A

Please choose the appropriate answer from the following options: A. RMSE B. SMSE C. TMSE D. MMSE

Which type of learning is most commonly used to create models on labelled continuous target?

A

Please choose the appropriate answer from the following options: A. Regression B. Classification C. Clustering D. Boosting

What is the type of function being used in logistic regression?

C

Please choose the appropriate answer from the following options: A. Sine function B. Multivariate Function C. Sigmoid function D. Linear function

In what type of machine learning labelled data is used

A

Please choose the appropriate answer from the following options: A. supervised learning B. reinforcement learning C. unsupervised learning D. active learning

What’s the name of the API that allows you to inspect the impact of convolutions on the images?

B

Please choose the appropriate answer from the following options: A. The model.convolu tions API B. The model.layers API C. The model.images API D. The model.pools API

Which parameter of the Image Generator is used to normalize the image?

B

Please choose the appropriate answer from the following options: A. normalize_ima ge B. rescale C. Rescale_image D. Normalize

Every Neuron in Feed Forward Neural Network has two parts viz. :

B

Please choose the appropriate answer from the following options: A. (1) A non- linear combination of inputs and (2) A linear activation function that converts the non- linear combination to another value. B. (1) Linear combination of inputs and (2) A non- linear activation function that converts the linear combination to another value. C. (1) A loss function and (2) An activation function D. (1) An activation function and (2) An optimizer function

If your training data is close to 1.000 accuracy, but your validation data isn’t, what’s the risk here?

C

Please choose the appropriate answer from the following options: A. You’re underfitting on your validation data B. No risk, that’s a great result C. You’re overfitting on your training data D. You’re overfitting on your validation data

Which method of Model object is used to extract the required layer from a NN model?

C

Please choose the appropriate answer from the following options: A. extract_layer() B. layer() C. get_layer() D. model_layer()

What is the shape of the following tensorflow object?

B

Please choose the appropriate answer from the following options: A. Tensorflow([2, 2,4]) B. Tensorflow([3, 3,4]) C. Tensorflow([2, 3,4]) D. Tensorflow([2, 9])

Which of the following is/are the loss functions used for linear regression?

D

Please choose the appropriate answer from the following options: A. mse B. mae C. mape D. All of these

Which of the following operations functions are available in tensorflow?

D

Please choose the appropriate answer from the following options: A. tf.pow() B. tf.sqrt() C. tf.exp() D. All of the above

How do you change the number of classes the model can classify when using transfer learning? (i.e. the original model handled 1000 classes, but yours handles just 2)

D

Please choose the appropriate answer from the following options: A. Ignore all the classes above yours (i.e. Numbers 2 onwards if I'm just classing 2) B. Use all classes but set their weights to 0 C. Use dropouts to eliminate the unwanted classes D. When you add your DNN at the bottom of the network, you specify your output layer with the number of classes you want

If my training data only has people facing left, but I want to classify people facing right, how would I avoid overfitting?

A

Please choose the appropriate answer from the following options: A. Use the ‘horizontal_fli p’ parameter B. Use the ‘flip’ parameter C. Use the ‘flip_vertical’ parameter around the Y axis D. Use the ‘flip’ parameter and set ‘horizontal’

The range of the output of tanh() function is

B

Please choose the appropriate answer from the following options: A. (0, 1) B. [-1, 1] C. (- ∞, ∞ ) D. [0,1]

Non-trainable parameters may appear in MLP architechure due to the addition of

BatchNormaliz -ation Layer

Please choose the appropriate answer from the following options: A. Sequential Layer B. BatchNormaliz ation Layer C. Conv2D Layer D. There is no concept of non trainable parameter in MLP

Aritifical Intelligence is about ________

B

Please choose the appropriate answer from the following options: A. Playing a game on computer B. Making an intelligent machine C. Programming on Machine with your own Intelligence D. Putting your intelligence in Machine

What is the job of an optimizer function?

C

Please choose the appropriate answer from the following options: A. Decides to stop training a neural network B. Figures out how to efficiently compile your code C. Generates a new and improved guess D. Measures how good the current guess is

How did you lock or freeze a layer from retraining?

D

Please choose the appropriate answer from the following options: A. tf.freeze(layer) B. tf.layer.frozen = true C. tf.layer.locked = true D. layer.trainable = false

Applying Convolutions on top of our Deep neural network will make training:

D

Please choose the appropriate answer from the following options: A. Stays the same B. Faster C. Slower D. It depends on many factors. It might make your training faster or slower, and a poorly designed Convolutional layer may even be less efficient than a plain DNN!

Which of the following is/are applications of Artifical Intelligence ?

D

Please choose the appropriate answer from the following options: A. Consumer Marketing B. Identificaiton technologies C. Machine Translation D. All of the above

Which parameter is used in fit() function to invoke callbacks?

D

Please choose the appropriate answer from the following options: A. callback= B. oncallback= C. oncallbacks= D. callbacks=

Which function is used to decide whether to take any action or not based on value generated by neurons?

D

Please choose the appropriate answer from the following options: A. Loss function B. Optimizer function C. Memory Function D. Activation function

How do you use Image Augmentation in TensorFLow?

D

Please choose the appropriate answer from the following options: A. With the keras.augment API B. With the tf.augment API C. You have to write a plugin to extend tf.layers D. Using parameters to the ImageDataGen erator

A.M Turing developed a technique for determining whether a computer could or could not demonstrate the artifical intelligence. Presently, this technique is called as

B

Please choose the appropriate answer from the following options: A. Algorithm B. Turing Test C. Boolean Algebra D. Linear Algebra

Why is transfer learning useful?

C

Please choose the appropriate answer from the following options: A. Because I can use all of the data from the original training set B. Because I can use all of the data from the original validation set C. Because I can use the features that were learned from large datasets that I may not have access to D. Because I can use the validation metadata from large datasets that I may not have access to

Which method is called when one epoch is completed?

C

Please choose the appropriate answer from the following options: A. on_end() B. on_training_c omplete() C. on_epoch_end () D. on_epoch_fini shed()

The job of softmax function is to

D

Please choose the appropriate answer from the following options: A. converge the algorithm B. make sure that the loss is minimum C. regularize the parameters D. convert the scores generated in the last hidden layer into probability values

What is the shape of the tensor given below: t2=tf.constant ([2,3,1,5])

B

Please choose the appropriate answer from the following options: A. (1, 4) B. (4,) C. 4 D. 5

________occu rs when the gradients become too large due to back- propagation.

A

Please choose the appropriate answer from the following options: A. Exploding Gradients B. Vanishing Gradients C. Long Short Term Memory Networks D. Gated Recurrent Unit Networks

Classically, topic models are introduced in the text analysis community for ____________ ____ topic discovery in a corpus of documents.

A

Please choose the appropriate answer from the following options: A. Unsupervised B. Supervised C. Semi- automated D. None of the above

Which of the following is/are true statement for advanced pre- processing topics in NLP? Statement 1: TF-IDF helps remove the outliers.

C

Please choose the appropriate answer from the following options: A. Statement 1 & 3 are True & Statement 2 is False B. Statement 2 & 3 are False & Statement 1 is True C. All the above Statements are True D. None of the above

N-grams are defined as the combination of N keywords together. How many bi-grams can be generated from a given sentence: “Hi Good Morning we welcome you all to the class”

C

Please choose the appropriate answer from the following options: A. 7 B. 8 C. 9 D. 10

In NLP, the process of removing words like “and”, “is”, “a”, “an”, “the” from a sentence is called as

C

Please choose the appropriate answer from the following options: A. Stemming B. Lemmatization C. Stop word D. All of the above

Bag of words in text preprocessing is a

B

Please choose the appropriate answer from the following options: A. Feature scaling technique B. Feature extraction technique C. Feature selection technique D. None of the above

One of the basic assumptions of LDA and LSA as a distributional hypothesis which means ____________ ______.

C

Please choose the appropriate answer from the following options: A. Different topics make use of similar words B. Similar topics make use of different words C. Similar topics make use of similar words D. None of the above

Deciding Insurance premium of a car based on online customers reviews is an application of ____________ _____.

C

Please choose the appropriate answer from the following options: A. Information Retrival B. Information Extraction C. Sentiment Analysis D. Text Summarizatio n

What is back propagation?

B

Please choose the appropriate answer from the following options: A. It is another name given to the curvy function in the perceptron B. It is the transmission of error back through the network to allow weights to be adjusted so that the network can learn C. It is the transmission of error back through the network to adjust the inputs D. None

Which of the following is/are Common uses of RNNs?

D

Please choose the appropriate answer from the following options: A. BusinessesHel p securities traders to generate analytic reports B. Detect fraudulent credit-card transaction C. Provide a caption for images D. All of the above

Which of the following is a loss function?

C

Please choose the appropriate answer from the following options: A. tanh B. ReLu C. Cross entropy D. Sigmoid function

How many trigrams phrases can be generated from the following sentence, after performing below text cleaning steps: *Stopword Removal *Replacing punctuations by a single space “#Unext- platform is a great source to learn @data_scienc e.”

B

Please choose the appropriate answer from the following options: A. 4 B. 5 C. 6 D. 7

Dissimilarity between words expressed using cosine similarity will have values significantly higher than 0.5

B

Please choose the appropriate answer from the following options: A. TRUE B. FALSE

The network that involves backward links from output to input is called

B

Please choose the appropriate answer from the following options: A. CNN B. RNN C. MLP D. Perceptron

Suppose you have a dataset from where you have to predict three classes. Then which of the following configuration you should use in the output layer?

B

Please choose the appropriate answer from the following options: A. Activation function = sigmoid, loss function = cross entropy B. Activation function = softmax, loss function = cross entropy C. Activation function = softmax, loss function = mean squared error D. Activation function = sigmoid, loss function = mean squared error

Which of the following statement is not correct?

It can only work for a single input and a single output

Please choose the appropriate answer from the following options: A. Neural networks mimic the human brain B. Neural networks can be used in image processing C. Neural networks can only work for a single input and a single output D. None

Which of the following activation function can not be used in the output layer of an image classification model?

A

Please choose the appropriate answer from the following options: A. ReLu B. Softmax C. Sigmoid D. None

Which of the following techniques can be used for keyword normalization in NLP, the process of converting a keyword into its base form?

A

Please choose the appropriate answer from the following options: A. Lemmatization B. Soundex C. CosineSimilarit y D. N-grams

What stemming refers to in text mining?

A

Please choose the appropriate answer from the following options: A. Reducing a word to its root B. Defining the parts of speech of a word C. Converting sentences to words D. None of the above

In text mining, how the words ‘lovely’ is converted to ‘love’

A

Please choose the appropriate answer from the following options: A. By stemming B. By tokenization C. By lemmatization D. None of the above

What is tokenization?

A

Please choose the appropriate answer from the following options: A. Breaking sentences into words B. Creating a set of vocabularies C. Removing stopwords D. All of the above

LSTM is a variation of-

D

Please choose the appropriate answer from the following options: A. Perceptron B. Multi Layer Perceptron Network C. Convolutional Neural Network D. Recurrent Neural Network

Which of the following neural networks is the best for machine translation?

B

Please choose the appropriate answer from the following options: A. 2D Convolutional Neural Network B. Recurrent Neural Network C. Multi Layer Perceptron Network D. None

___________ is a scoring of how rare the word is across documents.

A

Please choose the appropriate answer from the following options: A. Inverse Document frequency B. Term frequency C. File frequency D. None of the above