Data Science Technical Interview Questions

13. What do you understand by logistic regression? Explain one of

its use-cases.

Logistic regression is one of the most popular machine learning models used for
solving a binary classification problem, that is, a problem where the output can take
any one of the two possible values. Its equation is given by

Where X represents the feature variable, a,b are the coefficients, and Y is the
target variable. Usually, if the value of Y is greater than some threshold value, the
input variable is labeled with class A. Otherwise, it is labeled with class B.

14. How are univariate, bivariate, and multivariate analyses different

from each other?

Univariate Analysis Bivariate Analysis Multivariate Analysis

When only one variable When trends in two When more than two variables
is being analyzed variables are compared are considered for analysis to
through graphs like pie using graphs like scatter understand their correlations,
charts, the analysis is plots, the analysis of the the analysis is termed as
called univariate. bivariate type. multivariate.

15. What is K-means?

K-means clustering algorithm is an unsupervised machine learning algorithm that

classifies a dataset with n observations into k clusters. Each observation is labeled
to the cluster with the nearest mean.

16. How will you find the right K for K-means?

To find the optimal value for k, one can use the elbow method or the silhouette
method.

17. What do you understand by long and wide data formats?

In wide data format, you will find a column for each variable in the dataset. On the
other hand, in a long format, the dataset has a column for specific variable types &
a column for the values of those variables.
For example,

Wide Data Format

Long data format

Image Source: Mason John on Quora

18. What do you understand by feature vectors?

Feature vectors are the set of variables containing values describing each
observation’s characteristics in a dataset. These vectors serve as input vectors to a
machine learning model.
19. How does the use of dropout work as a regulariser for deep
neural networks?

Dropout is a regularisation method used for deep neural networks to train different
neural networks architectures on a given dataset. When the neural network is
trained on a dataset, a few layers of the architecture are randomly dropped out of
the network. This method introduces noise in the network by compelling nodes
within a layer to probabilistically take on more or less authority for the input values.
Thus, dropout makes the neural network model more robust by fixing the units of
other layers with the help of prior layers.

20. How beneficial is dropout regularisation in deep learning

models? Does it speed up or slow down the training process, and
why?

The dropout regularisation method mostly proves beneficial for cases where the
dataset is small, and a deep neural network is likely to overfit during training. The
computational factor has to be considered for large datasets, which may outweigh
the benefit of dropout regularisation.

The dropout regularisation method involves the random removal of a layer from a
deep neural network, which speeds up the training process.

21. How will you explain logistic regression to an economist,

physician-scientist, and biologist?

Logistic regression is one of the simplest machine learning algorithms. It is used to

predict the relationship between a categorical dependent variable and two or more
independent variables. The mathematical formula is given by

Where X is the independent variable, a,b are the coefficients, and Y is the
dependent variable that can take categorical values.

22. What is the benefit of batch normalization?

 The model is less sensitive to hyperparameter tuning.

 High learning rates become acceptable, which results in faster training of
the model.
 Weight initialization becomes an easy task.
 Using different non-linear activation functions becomes feasible.
 Deep neural networks are simplified because of batch normalization.
It introduces mild regularisation in the network.

23. What is multicollinearity, and how can you overcome it?

A single dependent variable depends on several independent variables in a

multiple regression model. When these independent variables are deduced to
possess high correlations with each other, the model is considered to reflect
multicollinearity.

One can overcome multicollinearity in their model by removing a few highly

correlated variables from the regression equation.

24. What do you understand by the trade-off between bias and

variance in Machine Learning? What is its significance?

The expected value of test-MSE (Mean Square Error, for a given value x0, can
always be decomposed into the sum of three fundamental quantities: the variance
of f0‘(x0), the squared bias of f0(x0), and the variance of the error terms e. That is,

E(y0 − f0‘(x0))2 = Var(f0‘(x0) + [Bias(f0‘(x0))]2 + Var(e)

Here the notation(y0 − f0(x0))2 defines the expected test MSE, and refers to the
average test MSE that one would obtain if they repeatedly estimated f using a large
number of training sets, and tested each at x0. Also, f0‘(x0) refers to the output of
the fitted ML model for a given input x0 and e is the deviation of the predicted
valuef0‘(x0) from the true value at a given x0.

The equation above suggests that we need to select a statistical learning method
that simultaneously achieves low variance and low bias to minimize the expected
test error. A good statistical learning method's good test set performance requires
low variance and low squared bias. This is referred to as a trade-off because it is
easy to obtain a method with extremely low bias but high variance (for instance, by
drawing a curve that passes through every single training observation) or a method
with a very low variance

but high bias (by fitting a horizontal line to the data). The challenge lies in finding a
method for which both the variance and the squared bias are low.

25. What do you understand by interpolating and extrapolating the

given data?

Interpolating the data means one is estimating the values in between two known
values of a variable from the dataset. On the other hand, extrapolating the data
means one is estimating the values that lie outside the range of a variable.
26. Do gradient descent methods always converge to the same
point?

No, gradient descent methods do not always converge to the same point because
they converge to a local minimum or a local optima point in some cases. It
depends a lot on the data one is dealing with and the initial values of the learning
parameter.

27. What is the difference between Supervised Learning and

Unsupervised Learning?

Supervised Learning Unsupervised Learning

If an algorithm learns something from If the algorithm does not learn anything
the training data so that the knowledge beforehand because there is no
can be applied to the test data, then it is response variable or training data, it is
referred to as Supervised Learning. referred to as unsupervised learning.

It is majorly used to make predictions for It is primarily used to perform analysis

a dependent variable. and group similar data points together.

Classification and Regression are an Clustering and dimensionality reduction

examples of Supervised Learning. are examples of unsupervised learning.

28. What is Regularization and what kind of problems does

regularization solve?

Regularization is basically a technique that is used to push or encourage the

coefficients of the machine learning model towards zero to reduce the over-fitting
problem. The general idea of regularization is to penalize complicated models by
adding an additional penalty to the loss function in order to generate a larger loss.
In this way, we can discourage the model from learning too many details and the
model is much more general.
There are two ways of assigning the additional penalty term to the loss function
giving rise to two types of regularization techniques. They are

 L2 Regularization
 L1 Regularization
In L2 Regularization, the penalty term is the sum of squares of the magnitude of
the model coefficients while in L1 Regularization, it is the sum of absolute values of
the model coefficients.

29. How can you overcome Overfitting?

We can overcome overfitting using one or more of the following techniques

1. Simplifying the model: We can reduce the overfitting of the model by reducing
the complexity of model. We can either remove layers or reduce the number of
neurons in the case of a deep learning model, or prefer a lesser order polynomial
model in case of regression.

2. Use Regularization: Regularization is the common technique used to remove

the complexity of the model by adding a penalty to the loss function. There are two
regularization techniques namely L1 and L2. L1 penalizes the sum of absolute
values of weight whereas L2 penalizes the sum of square values of weight. When
data is too complex to be modeled, the L2 technique is preferred and L1 is better if
the data to be modeled is quite simple. However, L2 is more commonly preferred.
3. Data Augmentation: Data augmentation is nothing but creating more data
samples using the existing set of data. For example, in the case of a convolutional
neural network, producing new images by flipping, rotation, scaling, changing
brightness of the existing set of images helps in increasing the dataset size and
reducing overfitting.
4. Early Stopping: Early stopping is a regularization technique that identifies the
point from where the training data leads to generalization error and begins to
overfit. The algorithm stops training the model at that point.
5. Feature reduction: If we have a small number of data samples with a large
number of features, we can prevent overfitting by selecting only the most important
features. We can use various techniques for this such as F-test, Forward
elimination, and Backward elimination.
6. Dropouts: In the case of neural networks, we can also randomly deactivate a
proportion of neurons in each layer. This technique is called dropout and it is a
form of regularization. However, when we use the dropout technique, we have to
train the data for more epochs.

30. Differentiate between Batch Gradient Descent, Mini-Batch

Gradient Descent, and Stochastic Gradient Descent.

Gradient descent is one of the most popular machine learning and deep learning
optimization algorithms used to update a learning model's parameters. There are 3
variants of gradient descent.
Batch Gradient Descent: Computation is carried out on the entire dataset in batch
gradient descent.
Stochastic Gradient Descent: Computation is carried over only one training
sample in stochastic gradient descent.
Mini Batch Gradient Descent: A small number/batch of training samples is used
for computation in mini-batch gradient descent.
For example, if a dataset has 1000 data points, then batch GD, will train on all the
1000 data points, Stochastic GD will train on only a single sample and the mini-
batch GD will consider a batch size of say100 data points and update the
parameters.