First Part

Question 1

What are the properties of normal distribution

Answer 1

Properties of Nornal Distribution:

Unimodal -one mode
Symmetrical -left and right halves are mirror images
Bell-shaped -maximum height (mode) at the mean
Mean, Mode, and Median are all located in the center
Asymptotic

Question 2

What is the goal of A/B testing

Answer 2

It is a statistical hypothesis testing for a randomized experiment with two variables A and B.

Question 3

What is sensitivity ,specificity ,accuracy and precision

Answer 3

Sensitivity or TPR(True Postive Rate)= TP/(TP+FN)
Specificity or TNR(True Negative Rate)= TN/(TN+FP)
Precision or PPV(Positive Predictive Value)=TP/(TP+FP)
ACC=(TP+TN)/(TP+FP+TN+FN)

Question 4

What is over-fitting

Answer 4

In over-fitting, a statistical model describes/follows the random error or noise instead of the underlying relationship. Over-fitting occurs when a model is excessively complex, such as having too many parameters relative to the number of observations. A model that has been over-fit has poor predictive performance, as it overreacts to minor fluctuations in the training data.

Question 5

What is under-fitting

Answer 5

Under-fitting occurs when a statistical model or machine learning algorithm cannot capture the underlying trend of the data. Under-fitting would occur, for example, when fitting a linear model to non-linear data. Such a model too would have poor predictive performance.

Question 6

What is Univariate analysis

Answer 6

Univariate analyses are descriptive statistical analysis techniques which can be differentiated based on the number of variables involved at a given point of time. For example, the pie charts of sales based on territory involve only one variable that can be referred to as univariate analysis.

Question 7

What are bivariate and multivariate analysis

Answer 7

Bi variate tries to see how two variables interact with each other and understand what are the differences between the two. Example is a scatter plot. Multi-Variate analysis is to do the same but with more than 2 variables

Question 8

What are eigen values and eigen vectors

Answer 8

Eigen Vectors are used for understanding linear transformations. In analysis they ae generally used for correlation or covariance matrix.Eigenvectors are the directions along which a particular linear transformation acts by flipping, compressing or stretching.

Eigenvalue can be referred to as the strength of the transformation in the direction of eigenvector or the factor by which the compression occurs.

Question 9

What is machine learning

Answer 9

Machine Learning explores the study and construction of algorithms that can learn from and make predictions on data.

Question 10

What is supervised learning

Answer 10

Supervised learning is the machine learning task of inferring a function from labeled training data. The training data consist of a set of training examples.
It essentially means that there is a target variable
Eg: Support Vector Machines, Regression, Naive Bayes, Decision Trees, K-nearest Neighbor Algorithm and Neural Networks

Question 11

What is unsupervised learning

Answer 11

Unsupervised learning is a type of machine learning algorithm used to draw inferences from datasets consisting of input data without labeled responses.
There is no target variable.
eg: Clustering, Anomaly Detection, Neural Networks and Latent Variable Models

Question 12

What is logistic regression

Answer 12

Logistic Regression often referred as logit model is a technique to predict the binary outcome from a linear combination of predictor variables.

Question 13

What is the logit model

Answer 13

Logit model=log(p/(1-p)) where p is the probability of the event occurring

Question 14

What are recommender systems

Answer 14

Recommender Systems are a subclass of information filtering systems that are meant to predict the preferences or ratings that a user would give to a product.

Question 15

What is collaborative filtering

Answer 15

The process of filtering used by most of the recommender systems to find patterns or information by collaborating viewpoints, various data sources and multiple agents.

Question 16

What do you mean by Deep Learning

Answer 16

Deep Learning is nothing but a paradigm of machine learning which has shown incredible promise in recent years. This is because of the fact that Deep Learning shows a great analogy with the functioning of the human brain.

Question 17

What are Artificial Neural Networks

Answer 17

Artificial Neural networks are a specific set of algorithms that have revolutionized machine learning. Neural Networks can adapt to changing input so the network generates the best possible result without needing to redesign the output criteria.

Question 18

What is Gradient Descent

Answer 18

A gradient measures how much the output of a function changes if you change the inputs a little bit. It simply measures the change in all weights with regard to the change in error. You can also think of a gradient as the slope of a function.
Gradient Descent can be thought of climbing down to the bottom of a valley, instead of climbing up a hill. This is because it is a minimization algorithm that minimizes a given function (Activation Function).

Question 19

What is an activation function. What are the various types of activation functions

Answer 19

Activation function A = “activated” if Y > threshold else not
Alternatively, A = 1 if y> threshold, 0 otherwise
1. Step Function
2.sigmoid function=1/(1+e^-x)
3. Tanh function
The tanh function is very similar to a scaled sigmoid function

Question 20

What is a tanh function

Answer 20

The tanh function is very similar to a scaled sigmoid function
(2/(1+e^-x))-1=2sigmoid(2x)-1

Question 21

How to solve gradient descent

Answer 21

It is based on the observation that if the multi-variable function F(x) is defined and differentiable in a neighborhood of a point a,-(diveregence)F(a).
When the function F is convex, all local minima are also global minima, so in this case gradient descent can converge to the global solution.

Question 22

What is Back Propagation and Explain it’s Working

Answer 22

Backpropagation is a training algorithm used for multilayer neural network. In this method, we move the error from an end of the network to all weights inside the network and thus allowing efficient computation of the gradient.

Question 23

What are the steps to back propagation

Answer 23

1. Forward Propagation of Training Data
2. Derivatives are computed using output and target
3. Back Propagate for computing derivative of error wrt output activation
4. Using previously calculated derivatives for output
5. Update the Weights

Question 24

What are the variants of Back Propagation?

Answer 24

Stochastic Gradient Descent: We use only single training example for calculation of gradient and update parameters.
Batch Gradient Descent: We calculate the gradient for the whole dataset and perform the update at each iteration.
Mini-batch Gradient Descent: It’s one of the most popular optimization algorithms. It’s a variant of Stochastic Gradient Descent and here instead of single training example, mini-batch of samples is used.

Question 25

What is the role of Activation Function?

Answer 25

The Activation function is used to introduce non-linearity into the neural network helping it to learn more complex function. Without which the neural network would be only able to learn linear function which is a linear combination of its input data. An activation function is a function in an artificial neuron that delivers an output based on inputs

Question 26

What is the formula on which decision tree is based on

Answer 26

For decision trees we use 2 methods
Impurity
Impurity is when we have a traces of one class division into other. This can arise due to following reason
1.We run out of available features to divide the class upon.
2.We tolerate some percentage of impurity (we stop further division) for faster performance. (There is always trade off between accuracy and performance).
For example in second case we may stop our division when we have x number of fewer number of elements left. This is also known as gini impurity.
Entropy
Entropy is degree of randomness of elements or in other words it is measure of impurity
Formula is