FAQ
- Is this only for people who plan to interview?
- No, it’s for everyone.
- Which type of interview does this book apply to?
- Quanitative trading, quantitative researching, Machine learning engineer, data science, software developer, then tyou will benefit from this.
- Who wrote this?
- Why is it called ’Data Science X’ if it encompasses machine learning, programming and quant finance?
- Because data science was the original name for computer science. Because machine learning is a subset of data science. And because Quant Finance is just an interesting application of Machine Learning and Math.
- How should I use this book?
- 1) For interview practice (duh). 2) As a textbook-like learning resource. 3) As a checklist of things to cover.
- Why are there so many Minecraft references?
- Mine your own business.
What is this?
- A website?
- A passion project?
- A textbook in the form of an interview?
- A repository of interview questions?
- Doubles as a HOME for all the sick materials I find, since I’m a bit of a hoarder.
- A diary of all the cool stuff that makes my brain tingle?
- The agglomeration of dozens of mock interviews, real interviews, catch-ups, events, courses, YouTube videos, books, and courses?
- All of the above.
Philosophies of this Book
1) 0% noise, 100% signal
The plan is to cut through the noise and provide pure signal for you, for the reader.
Quality does not increase with quantity of fancy words.
- Furthermore, most resources only tell you ’about’ models.
- They don’t really work through the exact mechanics of how they work
- and the ones that do, are so math-dense that they’re basically unreadable.
2) I’ve always believed asking is better than telling; that’s why this book is structured as an interview instead of an informatory textbook.
All the content you could ever want to cover is reframed as a question, either explicitly or implicitly.
3) I also believe in beautiful graphics - they make reading infinitely more exciting, and improve memory.
This book is designed to be fun and efficient.
Trust in first principles. Everything is just small layers of logic built on top of each other, everything. If you don’t get something, backtrack until you find the missing piece of logic
Structure of the Book
│
├── Act I (first major topic or unit)
│ ├── (1) (first section of Act I)
│ │ ├── (1.1) (first subsection of section (1))
│ │ │ ├── {A} (introductory question or problem)
│ │ │ ├── {B} (follow-up question with increased complexity)
│ │ │ ├── {C} (more advanced problem building upon the previous ones)
│ │ │ └── {D} (case study or application of the covered material)
│ │ │ ...
│ │ ...
│ ...
...
It is non-linear: (show each textbook as a node in a graph) and like a green yes/ No
Question Difficulty
All interview-style questions, with only 3 tiers of difficulty.
Within each section the questions get progressively harder, and sections may be partitioned into difficulty level.
We have tried hard to make the questions as close-ended as possible, because open-ended questions are annoying and unhelpful.
⍰ Easy/Hard questions target the concepts and theory.
☞ “What is 2+2?”
☞ “How do you understand feature importance in multivariate regression?”
⍰ Case studies are open-ended, hypothetical-situation, application-based questions and require high technical fluency. For example: ☞ “Whats the most fun way to make a snake game” (Jane Street, Final Interview) ☞ “How would you go about predicting churn for XYZ” (Atlassian, Final Interview) Some are hypothetical, some are real-world case studies.
Structure of a Question
Links are used for the sake of the reader, not just pretty obsidian graphs
This is an example question.
If the question is too long, it may overflow underneath. Questions can link to other questions as well.
The answer goes here.
###### ↳ This is a follow up question > ⚘ This is a helpful definition
☞ This will guide you through an answer… ∴ And reach a conclusion.
Your Knowledge Level
This book covers everything, but that doesn’t mean you have to cover everything.
It works independently of your level of knowledge… ☞ If you know nothing, you learn from the answers, ☞ If you know everything, you’ll get practice from the questions.
You’ll see what I mean.
The book is very loosely structured into only THREE main sections. Just read it from the start, in a linear fashion, and see how far you get. Questions are assorted. They will tend to get more niche, specialised and difficult. There are brain teasers sprinkled throughout. more of a repository than a textbook; but it works if you read it top to bottom. It’s inherently non linear; an non-directed graph, as you can see. Jump around from topic to topic and slowly fill in the gaps. If you have enough knowledge to read it linearly, you probably hold 3 different PhD’s. I have used tags and structuring to allow you to hop between different topics. Questions are designed to be MECE
How to Read
- How would you feel if Python only executed 60% of your code?
- All unnecessary noise, cleaned away (sketch)
- technically the book is self contained… you can…
- think of scary the word ‘cardiovascular’ looked the first time you read it. Simple logic hides behind all the scary words in this book. It’s called Jargon.
- Understanding > Memorising > Reading
- Look say cover write check. Just because you read the answer doesn’t mean you will answer it in an interview.
- In your own words - or even just reading from the answers. Your interview cannot be the first time you’ve ever tried to articulate these concepts aloud. Explain them to your uncaring friend, whatever it takes.
- Follow the pointers ☞ (like it’s
C++).