BC DS

10 machine learning questions you should be able to answer.

I started experimenting with a questionnaire before interviewing candidates. Applicants submit a video or written answers to apply for the position.

Thread: Let's look at 10 of these questions.

The questionnaire solves a couple of problems:

1. Weeds out candidates that aren't ready for the job.

2. It significantly cuts down the time we spend talking synchronously.

One problem: It requires extra time from the candidate before the phone interview.
Here are some numbers on a job posting before using the questionnaire:

• 249 applicants
• 13 good applications

That's only 5%.

This sucks.
Numbers after using the questionnaire:

• 44 applicants
• 21 good applications

That's 47%.

This is much better.
I look at it this way:

Sooner or later, I'll ask these questions, and the candidate will have to answer them.

I'd rather give them space and the opportunity to think about their answers without any pressure.

Those who can't or don't want to answer won't apply. That's fine.
I'm not worried about candidates searching for answers or asking someone else.

They will fool the questionnaire, not the interview that will come later.

Remember:

• The goal of the questionnaire and the resume is just to decide who moves to an interview.
Questions must be open-ended. No gotchas.

I want to allow candidates to think creatively and elaborate on their answers.

Preferred: A video response with the answers. Fallback is written answers if the candidate is not comfortable with video.
There are 10 questions on the questionnaire.

These cover basic principles that we want to ensure candidates understand.

Every candidate receives the same questions.

Let's go over them:
1. Provide a couple of examples where you used Supervised and Unsupervised machine learning techniques.

2. Walk us through the process you follow to select the best machine learning algorithm to solve a problem.

3. Explain the trade-off between bias and variance.
4. How do you decide how you should split your dataset when working in a supervised machine learning problem?

5. How do you handle imbalanced datasets in a classification problem?

6. Explain how you would approach reducing overfitting on a model.
7. Explain three different problems where using accuracy, precision, and recall is the best metric, respectively.

8. What would you expect to see as you vary the batch size when training a neural network?
9. What would you expect to see as you vary the learning rate when training a neural network?

10. What are the advantages of a Convolutional Neural Network over a fully connected network for image classification?
I post threads like this every week. You can find them here: @svpino.

Follow along for a good bunch of practical tips and epic stories about my experience with machine learning.

More from Santiago

More from Ds

1/

Get a cup of coffee.

In this thread, I'll walk you through 2 probability concepts: Standard Deviation (SD) and Mean Absolute Deviation (MAD).

This will give you insight into Fat Tails -- which are super useful in investing and in many other fields.


2/

Recently, I watched 2 probability "mini-lectures" on YouTube by Nassim Taleb.

One ~10 min lecture covered SD and MAD. The other ~6 min lecture covered Fat Tails.

In these ~16 mins, @nntaleb shared so many useful nuggets that I had to write this thread to unpack them.

3/

For those curious, here are the YouTube links to the lectures:

SD and MAD (~10 min):
https://t.co/0TwubymdE6

Fat Tails (~6 min):

4/

The first thing to understand is the concept of a Random Variable.

In essence, a Random Variable is a number that depends on a random event.

For example, when we roll a die, we get a Random Variable -- a number from the set {1, 2, 3, 4, 5, 6}.

5/

Every Random Variable has a Probability Distribution.

This tells us all the possible values the Random Variable can take, and their respective probabilities.

For example, when we roll a fair die, we get a Random Variable with this Probability Distribution:

You May Also Like

Trump is gonna let the Mueller investigation end all on it's own. It's obvious. All the hysteria of the past 2 weeks about his supposed impending firing of Mueller was a distraction. He was never going to fire Mueller and he's not going to


Mueller's officially end his investigation all on his own and he's gonna say he found no evidence of Trump campaign/Russian collusion during the 2016 election.

Democrats & DNC Media are going to LITERALLY have nothing coherent to say in response to that.

Mueller's team was 100% partisan.

That's why it's brilliant. NOBODY will be able to claim this team of partisan Democrats didn't go the EXTRA 20 MILES looking for ANY evidence they could find of Trump campaign/Russian collusion during the 2016 election

They looked high.

They looked low.

They looked underneath every rock, behind every tree, into every bush.

And they found...NOTHING.

Those saying Mueller will file obstruction charges against Trump: laughable.

What documents did Trump tell the Mueller team it couldn't have? What witnesses were withheld and never interviewed?

THERE WEREN'T ANY.

Mueller got full 100% cooperation as the record will show.