# MakeMyTrip Interview for Data Engineer Q.1. What is Regression ? What is Classification ?
Ans. Regression : Target variable continous.
Classification : Target variable is discrete

Q.2. What are the error metrics used for each of them ?
Ans. Regression : SSE
Classification : Confusion metrics

Q.three. Why Accuracy doesn’t assist when there’s a class imbalance ?
Ans. Because accuracy will get drawn by the bulk class.

Q.four. How to deal with class imbalance ?
Ans. 1. Use efficiency metrics as Area Under ROC Curve
2. Penalize Algorithms
three. Use Tree-Based Algorithms like RF, Gradient Boosted Trees

Q.5. What is knn ?
Q.6. What is k-means ?
Explanation right here : https://www.quora.com/How-is-the-k-nearest-neighbor-algorithm-different-from-k-means-clustering

Q.7. y=ax+b is a linear mannequin. Can you inform me if y=ax^2 + bx + c can also be a linear mannequin ?
Ans. y=ax^2 + bx + c can also be linear as x^2 can be represented as X.
So, the precise relationship won’t be linear however the mannequin fitted is linear

Q.eight. What is SSE and RMSE ? Why to make use of RMSE and never SSE ?
Ans. RMSE has imply worth however SSE is whole worth.

Q.9. Does a low RMSE denote overfitting ?

Q.10. How to resolve overfitting ?
Ans. 1. Cross-validation
2. Regularization
three. Ensembling

Q.11. Why knn shouldn’t be a mannequin ?
Ans.11. It is a lazy mannequin.

Q.12. Write code in Python for the next issues :
(a)
bookings desk :
id, date, platform
1, 12/three, android
2, 12/three, ios
three, 13/three, android
four, 13/three, ios
5, 13/three, android
6, 14/three, ios
7, 14/three, android
For every date, what number of bookings are from android and what number of from ios ?
df1.groupby([‘date’, ‘platform’]).rely()

(b)
knowledge = [‘cat’, ‘bat’, ‘rat’, ‘cat’, ‘rat’]
Give the rely of every distinctive aspect of the listing
import pandas as pd
knowledge = [‘cat’, ‘bat’, ‘rat’, ‘cat’, ‘rat’]
df = pd.DataBody(knowledge, columns=[‘Category’])
vc = df[‘Category’].value_counts()

Join Free Interview Preparation Course

If you want GeeksforGeeks and want to contribute, it’s also possible to write an article utilizing contribute.geeksforgeeks.org or mail your article to [email protected] See your article showing on the GeeksforGeeks fundamental web page and assist different Geeks.

Please Improve this text in the event you discover something incorrect by clicking on the “Improve Article” button under.

Article Tags :

thumb_up
Be the First to upvote.

Please write to us at [email protected] to report any subject with the above content material.