In the first case study, the query which tries to generate the ordered count of events for last week have incorrect query syntax. It doesn't yields the expected result. I want to submit the new query but I am not able to create a branch in this repo in order to create a pull request.

As written in Wikipedia, ROC stands for TPR against FPR.

But in the theory.md, it says "The diagrammatic representation that shows the contrast between true positive rate vs true negative rate.", I think may be it's a little mistake.

Regards,

Good works!

I think we may create table of contents for markdown files of questions.
(I can submit a pull request for this if you don’t mind 😄 )

e.g.,

theory.md

Table of contents

• Supervised machine learning
• Linear regression
• Validation
• Classification
• Regularization
• Feature selection
• Decision trees
• Random forest
• Gradient boosting
• Parameter tuning
• Neural networks
• Optimization in neural networks
• Neural networks for computer vision
• Text classification
• Clustering
• Dimensionality reduction
• Ranking and search
• Recommender systems
• Time series

technical.md

Table of contents

• SQL
• Coding (Python)
• Algorithmic Questions

Would be nice to have the answers to questions collapsed initially, so that reader can think a bit about the answer first, then it can be expanded to see the answer.
I would be happy to make a PR with such change. I am trying to understand how can i build site locally before pushing changes to this repo.
Thanks!

Q4 standard deviation code error

The Error was in line.

avg = mean(numbers)

Mean can be imported from statistics and not math module.
Added the following code to correct it.

from statistics import mean

A pull request is already generated.

I read your article in HackerNoon and came to this repo.
That was an interesting and very helpful post.

Will you also share the database.
If you do not have a table schema, you can also post the CSV files.

select * from table1

create a CSV file and then you can share on Github.

Hello, There is kind of an advance version of this Repo here https://github.com/kojino/120-Data-Science-Interview-Questions.
I would recommend you to talk to the owner and merge them

Thanks!

Visualization/Images will make this repo a one-stop destination for all interview prep.
Is there any reason those are not added??
I will be willing to make it more appealing to the Eye.

Most of the python code(def) in technical.md is not called.As a result one has to copy the example from markdown examples and call the function to understand its working.
In some cases the markdown example is insufficient to understand the code.
For example :
9) Counter. We have a list with identifiers of form “id-SITE”. Calculate how many ids we have per site.

This Question is very brief and the markdown example guides on what we are trying to do but does not specify how is the input supplied into the function.

The fibonacci4 is defined as follow:

def fibonacci4(n):
    '''Top down + memorization (dictionary), complexity = O(n)'''
    dic = {1:1, 2:2}
    if n not in dic:
        dic[n] = fibonacci4(n-1) + fibonacci4(n-2)
    return dic[n]

Since the dic is a local variable, it is not shared during the recursion (you can check it by printing at each step)

A way to take advantage of memoization is to use the package functools with the fibonnaci1 implementation:

from functools import lru_cache

@lru_cache()
def fibonacci1(n):
    if n == 0 or n == 1:
        return n
    else:
        return fibonacci1(n - 1) + fibonacci1(n - 2)

You can find a quick comparison in the following image:

AB testing is asked in almost all interviews. There are many intricacies which show an interviewer if there is a deep understanding or not. Adding some of these questions would be very helpful.

We can remove lines of codes in the 6th point: "remove duplicates".
This is a patch: simotrone@43b31f7

This will make it easier to try to work out the problems before seeing the answer

2 + 2 = 4

I can implement this if it is a feature people want