Gain insights into ML system design, state-of-the-art techniques, and best practices for scalable production. Learn from top researchers and stand out in your next ML interview.

Machine Learning System Design is an important component of any ML interview. The ability to address problems, identify requirements, and discuss tradeoffs helps you stand out among hundreds of other candidates. Readers of this course able to get offers from Snapchat, Facebook, Coupang, Stitchfix and LinkedIn. 

This course will help you understand the state of the practice on model techniques along with best practices in applying ML models in production at scale. Once you finished the course you can learn more use-cases at: http://mlengineer.io/

Once you're done with the course, you will be able to apply and leverage knowledge from top researchers at tech companies. You will have up to date knowledge in model techniques from hundreds of the latest research and industry papers. There is even a chance that the interviewers will be surprised at the depth of your knowledge.

I really found the quizzes very helpful for testing my ML understanding. Also, the resources shared helped me a lot for revising concepts for my interview preparation. This course will definitely help engineers crack Machine Learning Engineering and Data Science interviews

Senior Data Scientist at Amazon

I have been using your github repo to prep for my interviews and got an offer with NVIDIA with their data science team. Thanks again for your help!

Data Scientist at NVIDIA

I really like what you've built, it'll help a lot of engineers.

MLE at Facebook

It's well organized and the illustrations are well done. Being able to visualize and walk through the steps in order is really helpful in system design. The hints for quizzes is a nice addition, a hint is given if you get stuck in a real interview, mimicking how a real interview 

DS at Fortune 500

Andrew

I just heard back from the recruiter that I passed the Google L5 HC. Thank you very much for sharing the resources on GitHub and for the course on educative.io!

Google Machine Learning Engineer, L5

I got the offer from Intuit. Thanks so much, it would not have been possible without your help.

Senior Machine Learning Engineer, Intuit

I got Google, Facebook, Apple, Tesla, Cruise offer for Senior ML engineer. I thought the course is super helpful. 

Senior Machine Learning Engineer at Cruise

Machine Learning System Design

# 3. Model

## Feature engineering
|**Features**   |**Feature engineering**   |**Description**   |
|:-: |:-: |:-: |
|User profile: job title, industry, demographic, etc. |For low cardinality: Use one hot encoding. Higher cardinality: use Embedding.   |   |
|Connection strength between users   |   |Represented by the similarity between users. We can also use Embedding for users and measure the distance vector.   |
|Age of activity   |Considered as a continuous feature or a binning value depending on the sensitivity of the Click target.    |   |
|Activity features   |Type of activity, hashtag, media, etc. Use Activity Embedding and measure the similarity between activity and user.  | |
|Cross features   |Combine multiple features.  |See the example in the Machine Learning System Design Primer. Read about [cross features](https://www.educative.io/courses/machine-learning-system-design/feature-selection-and-feature-engineering#3-conjunction) |


## Training data

Before building any ML models we need to collect training data. The goal is to collect data across different types of posts, while simultaneously improving user experience. Below are some of the ways we can collect training data:

  - Rank by chronicle order: This approach ranks each post in chronological order. Use this approach to collect click/not-click data. The trade-off here is serving bias because of the user’s attention on the first few posts. Also, there is a data sparsity problem because different activities, such as job changes, rarely happen compared to other activities on LinkedIn.
  - Random serving: This approach ranks post by random order. This may lead to a bad user experience. It also does not help with sparsity, as there is a lack of training data about rare activities. 
  - Use a Feed Ranking algorithm: This would rank the top feeds. Within the top feeds, you would permute randomly. Then, use the clicks for data collection. This approach provides some randomness and is helpful for models to learn and explore more activities.

Based on this analysis, we will use an algorithm to generate training data so that we can later train a machine learning model.

We can start to use data for training by selecting a period of data: last month, last 6 months, etc. In practice, we want to find a balance between training time and model accuracy. We also downsample the negative data to handle the imbalanced data. 

## Model

### Selection
- We can use a probabilistic sparse linear classifier (logistic regression). This is a popular method because of the computation efficiency that allows it to work well with sparse features.

- With the large volume of data, we need to use distributed training: Logistic Regression in Spark or Alternating Direction Method of Multipliers.

- We can also use deep learning in distributed settings. We can start with the fully connected layers with the Sigmoid activation function applied to the final layer. Because the CTR is usually very small (less than 1%), we would need to resample the training data set to make the data less imbalanced. It’s important to leave the validation set and test set intact to have accurate estimations about model performance.

Learn about the Feed Ranking system architecture and the model requirements.

Machine Learning Primer

Video Recommendation

Feed Ranking

Ad Click Prediction

Rental Search Ranking

Estimate Food Delivery Time

Conclusion

Machine Learning Knowledge

Machine Learning Model Diagnosis

Feed Ranking Model

3. Model

Feature engineering

Features	Feature engineering	Description
User profile: job title, industry, demographic, etc.	For low cardinality: Use one hot encoding. Higher cardinality: use Embedding.
Connection strength between users		Represented by the similarity between users. We can also use Embedding for users and measure the distance vector.
Age of activity	Considered as a continuous feature or a binning value depending on the sensitivity of the Click target.
Activity features	Type of activity,