Reasons for my Rejections in Data Science Interviews

Image source: Google Image

My warm welcome to all the readers!

Few reasons why I got rejected in Data Science interviews and what I learned from my rejections. This might help you to get yourself ready before the interview. Please do share your interview experiences.

1st interview:

1. They might even ask if we have gone through any research paper lately to see how enthusiastic we are to update ourselves. I didn’t go through any research paper then. So it is better to find time to go through research papers.
2. I told xyz is my favorite algorithm but failed to explain in detail about it. Have a good idea of other concepts and deep knowledge of our favorite concepts.

2nd interview:

I was notified about the interview just 3 hours before. I was not prepared for it at all.

1. We should never think if we apply for the job now, they will schedule after 3–4 days and we can prepare in these 3–4 days. They can schedule at any time.
2. We should always be prepared. At least have an idea of what we have learned so far. This is better than saying I have not revised anything yet.
3. Then, I was not prepared well. I kept saying for a few questions as ‘I know but I have to revise them.’

3rd interview:

I spent 4 days on revision and I was well prepared with the concepts this time. But the whole game turned to reverse. Not a single question was asked from the concepts. They asked only about my projects and this time I was not prepared with it.

1. I managed only to tell the objective, what is it about, and which algorithm I used in 2 minutes hardly. This was the reason for my rejection.

We should explain our projects in detail like

• what is the project all about and objective?
• What analysis did we make from the EDA?
• Which algorithm(s) we used and why? What are the hyperparameters we tuned and why?
• What performance metric we used and why? Why not another metric?
• What feature engineering we applied and how it helped in our model performance?
• In the case of deep learning, how many layers we used, and how we found only those many layers are good for the model performance?
• Explain the whole thing about the architecture like which activation function and why? Why dropout layers? Which pool layer and why only that and why not another pool layer? Which loss function and why only that?

In short, for everything we did, we should explain what we did, how we did, why we did only that, and why not an alternative method.

At the end of the day, we should never feel low with rejections. We should always keep ourselves up no matter what. We should know for sure that rejections build a strong base only if we take it positively. Sooner or later but surely we will embrace our dreams tightly.

Please share your interview experiences and let us keep motivating each other.

Connect me:

1. LinkedIn: https://www.linkedin.com/in/sandeep-panchal-682734111/
2. GitHub: https://github.com/Sandeep-Panchal

Thank you all for reading this blog.