Quinbay Data Scientist Interview Questions and Answers

Q1. How do you narrate a story using data, and what specific factors do you prioritize during this process?
Ans. 
Data storytelling involves transforming data insights into a compelling narrative to drive understanding and action.
• Identify the audience: Tailor the narrative to the knowledge level and interests of the audience, e.g., technical vs. non-technical stakeholders.

• Define the key message: Focus on the main takeaway you want the audience to remember, such as the impact of a marketing campaign on sales.

• Use visuals effectively...read more
Answered by AI
Add your answer
Q2. How proficient are you in using Python for data science, and can you provide specific examples of your experience?
Add your answer

Q1. About CNN architecture and how it is relevant for textual data
Ans. 
CNNs can effectively process textual data by capturing spatial hierarchies and local patterns in text.
• CNNs use convolutional layers to extract features from text, similar to how they process images.

• They can capture n-grams (e.g., phrases) by applying filters of varying sizes across the text.

• Pooling layers help in reducing dimensionality while retaining important features, making the model more efficient.

• Example: Text c...read more
Answered by AI
Add your answer
Q2. Questions related to probability therom
Add your answer

Q1. Reverse a string in list
Ans. 
Reverse a string in a list of strings
• Iterate through each string in the list

• Use the built-in function to reverse each string

• Store the reversed strings in a new list
Answered by AI
Add your answer
Q2. Explain joins in sql
Ans. 
Joins in SQL are used to combine rows from two or more tables based on a related column between them.
• Joins are used to retrieve data from multiple tables based on a related column

• Types of joins include INNER JOIN, LEFT JOIN, RIGHT JOIN, and FULL JOIN

• Example: SELECT * FROM table1 INNER JOIN table2 ON table1.column = table2.column
Answered by AI
Add your answer

I was given assigment on a simple problem where task was to analyse and create a working solution for a problem statement

Q1. What is bert algorithm? How does it work.
Ans. 
BERT (Bidirectional Encoder Representations from Transformers) is a pre-trained natural language processing model.
• BERT is a transformer-based machine learning algorithm developed by Google.

• It is designed to understand the context of words in a sentence by considering both the left and right context simultaneously.

• BERT has been pre-trained on a large corpus of text data and can be fine-tuned for specific NLP tasks like ...read more
Answered by AI
Add your answer
Q2. Can you explain regression in logistic regression?
Ans. 
Logistic regression is a type of regression analysis used to predict the probability of a binary outcome.
• Logistic regression is used when the dependent variable is binary (e.g. 0 or 1, yes or no).

• It estimates the probability that a given input belongs to a certain category.

• The output of logistic regression is transformed using a sigmoid function to ensure it falls between 0 and 1.

• It uses the logistic function to model ...read more
Answered by AI
Add your answer

Q1. Bias variance trade off
Add your answer
Q2. What is AB testing
Ans. 
AB testing is a method used to compare two versions of a webpage or app to determine which one performs better.
• AB testing involves creating two versions (A and B) of a webpage or app with one differing element

• Users are randomly assigned to either version A or B to measure performance metrics

• The version that performs better in terms of the desired outcome is selected for implementation

• Example: Testing two different call...read more
Answered by AI
Add your answer
Q3. Basic traditional ML question about ML metrics, bagging boosting etc.
Add your answer

It was a classification problem

Q1. Questions about assignment
Add your answer
Q2. Questions from resume.
Add your answer
Q3. Questions based on probability, statistics and loss functions
Add your answer

Q1. Tell me about yourself?
Add your answer
Q2. Tell me about your recent Project in Detail?
Add your answer
Q3. What is the Diff between Parameters and Hyper Parameters?
Ans. 
Parameters are learned from data; hyperparameters are set before training to control the learning process.
• Parameters are internal to the model, like weights in a neural network.

• Hyperparameters are external configurations, such as learning rate or number of trees in a random forest.

• Example of parameters: weights in linear regression.

• Example of hyperparameters: batch size, number of epochs in training.
Answered by AI
Add your answer
Q4. Can we use Logistic Regression for Multi Class Classification?
Ans. 
Yes, Logistic Regression can be adapted for multi-class classification using techniques like One-vs-Rest or Softmax regression.
• Logistic Regression is inherently binary, but can be extended to multi-class using One-vs-Rest (OvR) strategy.

• In OvR, a separate binary classifier is trained for each class, treating it as the positive class and all others as negative.

• Another approach is Softmax regression, which generalizes lo...read more
Answered by AI
Add your answer
Q5. What will happen if we will increase the value of K in KNN?
Ans. 
Increasing K in KNN can lead to smoother decision boundaries but may also introduce bias and reduce model sensitivity.
• Higher K values can smooth out noise in the data, leading to more generalized predictions.

• For example, with K=1, the model may overfit to noise, while K=10 may provide a more stable prediction.

• Increasing K can lead to underfitting, where the model fails to capture the underlying patterns in the data.

• Cho...read more
Answered by AI
Add your answer
Q6. If you have 50 GB of training data and you want to train your Neural Network on you Local 2 GB RAM, what will you do?
Ans. 
Use techniques like data sampling, mini-batch training, or cloud resources to handle large datasets on limited RAM.
• Data Sampling: Use a subset of the data, e.g., 5 GB, to train the model initially.

• Mini-Batch Training: Train the model on smaller batches of data, e.g., 256 MB at a time.

• Data Augmentation: Generate synthetic data to reduce reliance on the full dataset.

• Use Cloud Services: Leverage platforms like AWS or Goog...read more
Answered by AI
Add your answer
Q7. What is imbalanced Data?
Add your answer

Basic aptitude question

Q1. Basic analysis and ML questions , more focused on approach instead of coding
Add your answer

Q1. As like other HR rounds
Add your answer