Skip to content

analyticsbot/ml-system-design

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

8 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

ML Design Interview Guide

Overview

An ML Design Interview assesses your understanding of designing machine learning systems and solutions. Interviewers look for ability to ask questions, your approach to problem-solving, understanding of ML algorithms (and more importantly the ability to compare and pick multiple ML algorithms based on data, problem, and other variables at hand), and ability to scale solutions. You will likely be asked to design a model or an end-to-end system based on a real-world problem.

Steps to Answer an ML Design Question

  1. Problem Understanding/ Clarifying Requirements: Clearly define the problem, ask clarifying questions, and confirm the scope.
  2. Frame as ML Problem: How can the business problem be converted to an ML problem.
  3. Data Prep, Feature Engineering and Dataset creation: Outline the data (start with high level sources such as user, product, etc), create features from the data, and talk about volume of data required. Also, mention how will you create the train, val, test set.
  4. Model Selection: Choose appropriate models or algorithms for the problem. Important to compare more than
  5. System Architecture: Design the entire ML pipeline, including data ingestion, training, validation, and serving.
  6. Evaluation Metrics: Define metrics for evaluating model performance.
  7. Scalability and Deployment: Discuss how you’ll handle scaling, deployment, monitoring, and retraining.
  8. Risks & Limitations: Identify potential pitfalls and limitations in your design.

Resources

Please browse through the directory 'files' and it will contain the documents for multiple ML system design interviews.

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors