 ## What questions will you get in a Data Science and Machine Learning job interviews?

In this blog, I will acquaint you with some of the most frequently asked question in data science and Machine learning interviews.

1. What does data science stand for? Differentiate between supervised and unsupervised learning?
• Data science is the area of study in which domain experience, coding expertise; math and statistics skills are combined to extract practical data insights. This makes us wonder how is this different to what statisticians have been doing?

This can be answered if we understanding the difference between explaining and predicting. Following are the differences between supervised and unsupervised learning.

Check out the Data Science Masters Programme @ Fingertips

 Supervised learning Unsupervised learning Uses a training data set Uses the input data set Input data is labeled Input data is not labeled Enables classification and regression Enables classification, Dimension Reduction and Density Estimation. Used for prediction Used for analysis

1. What is the difference between KNN and K-Means clustering
• K-Means clustering is an unsupervised learning based algorithm, whereas K-Nearest Neighbors is a supervised learning classification algorithm. Although the methodology seems similar, in K-Nearest Neighbors you need labeled data so you can classify and unlabeled point into it(Hence, the nearest neighbor part). Only a set of unlabeled points and a threshold are necessary for K-Means clustering: the algorithm will accept unlabeled points and end up learning how to cluster them into groups by computing the distance mean between different points.
• The crucial difference is, that K-Means clustering is unsupervised learning based therefore doesn’t require labeled data points, whereas KNN does –and hence its supervised learning based.

1. Explain precision and recall
• Recall is also called true positive rate: the amount of positives your model says compared to the actual number of positives in the data. Precision is also regarded as the positive predictive value, and compared to the number of positives it actually states, it is a measure of the amount of precise positives that the model states.

1. Differentiate between generative and discriminative model

A generative model learns various categories of data whereas a discriminative model will simply learn the distinction between various data categories. As far as the classification tasks are concerned discriminative models usually outperform generative models.

1. What does F1 score mean? How to use it?
• The f1 score is a calculator of performance. It is a weighted average of the accuracy and recall of a model, with the best outcomes appearing to be 1 and the worst ones appearing to be 0.

1. What is a hash table?
• Hash table is a type of data structure that generates an associative array. By using a hash function, a key is mapped to some value. They are also used for tasks like indexing of databases.

1. List the data types supported by JSON
• The six rudimentary JSON datatypes that you can manipulate are – Number, string, objects, arrays, Booleans, and Null values.
##### Tags: 