Dec 2022 Newsletter 2022 — Dec & Nov Q4

Data Science Bootcamp


11 min readDec 27, 2022


Welcome! And happy holidays to all the developers around the world!!

A REMINDER: Welcome new paid subscribers! Please visit and create an account so we can activate your pro profile!! We have made as many flash cards free for the time being.

Terms of service By using our products, contents, any channel, you agree with our terms of service, privacy policies and any legal documents. If you disagree, please immediately stop using. Thank you. The legal docs can be found on the payment page. Our terms of service might have changed since the last time you saw it. Please be sure to review again.

Developer lifestyle is good

Layoffs are horrible. Despite the terror unleashed in tech, there’s still an impressive developer lifestyle to look forward to: from Adobe holiday party with surprise appearance of Katy Perry to OpenAI’s beautiful San Francisco office hosting young female developers wine and dine while explaining DALLE concepts and handing out luxury picnic blankets, to Solana’s free co-working and hackathon opportunities. In our next newsletter, we will discuss how to handle layoffs and the psychology, and why being laid off is not your (or your friend/ family) fault. Disclaimer : not a career counselor.

LIFESTYLE YOUTUBE VIDEOS — more coming in the next newsletter

Follow us on tiktok

Easter Eggs and Pro Tips are for paid members only. Other flashcards are generally available.

Uniqtech Data Science Bootcamp is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.


Andrew Ng (founder of Coursera) ’s free book on how to break into AI, check it out.

EASTER EGGS [high quality, pro, paid members only, best resources this month]

How do the best companies handle AI research and product design? Tesla [pro, recruiting] (scroll down to see more Tesla flashcards), an important foundational model at OpenAI [pro] that you should know and it is not GPT-3, ByteDance (Tiktok!) — Deep Learning at Tiktok [pro]. Tech stack at Random but important did you know we just mentioned two companies founded by Elon Musk? OpenAI and Tesla!

Career Easter Eggs: study tips [pro] from Coursera founder and Stanford professor Andrew Ng. Living with ADHD [comic].

Disclaimer: This is not career advice. We do not have a career counselor on staff.

Talking about GPT-3, both ChatGPT and DALLE-like generative arts are trending right now. Everyone is posting their AI selfie on instagram The holiday hero image of this newsletter is also generated using prompted AI. Scroll down to read about the model, magic behind generative arts — Stable Diffusion.

Have you seen a lot of glamorous digital avatars popping up on your friends social media profiles? Chances are these are generative art created using LensaAI which utilizes Stable Diffusion. See our stable diffusion flash cards below.

When GPT-3 first came out, many experimented with its creative and imaginative writing capabilities. Now ChatGPT made it even easier to interact with. Creating with AI can help us explore creative ideas and iterate quickly. It can surcharge artist and non-artist creativity. Ask it to write a rap about machine learning and let us know what you get! While GPT writing can still give hints that it is synthetic, sometimes it is hard to distinguish AI writing from human writing. One college student used GPT-3 to generate a blog that caused quite some stir on reddit. You can now also use AI to generate stock images — GhostlyStock and use AI to generate resumes e.g. (mentioning a technology is not an endorsement).

People have been obsessed with generating digital avatars with Lensa Ai and Astria Tailor-made Al image generator (Disclaimer: mentioning a tech is not an endorsement).

Learn more about these new generative trends by checking out our GPT, OpenAI and GAN flash cards.

PS: While we are still working on our newseltter text-embedding-ada-002 is already out on Dec 15, 2022! Previous DaVinci was the smart model. OpenAI is on a roll.

Uniqtech Data Science Bootcamp is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.


The basics of statistical analysis : calculate R Squared, correlation [definition], visualize R Squared. Improve your skills as a developer — learn about the best development tools that data scientists use. Talking about data scientists, what are the most popular programming language among data scientists? The survey has spoken. Here’s a resource book for data scientists — Data science book resources [pro]. Kaggle is a famous data science competition site, here’s how to make the most of it. [pro]

Linear Algebra

Calculating a common distance metric — L2 Norm ||u|| calculating norm, distance [public, linear algebra, dot product].


What is normalization normalizing (definition) SQL, relational database. What is normalizing? (Database, SQL) 02 MongoDB continues to be an important and relevant database application. Here are some pro tips about MongoDB.

Heatmap is a great data visualization tool. Here’re some [pro] tips about Heatmap. Data visualization for confusion matrix.

Understanding confusion matrix, precision, recall, sensitivity, specificity [formula, visualization, pro tip].

Basic R Programming installation and rstudio.

Gaussian distribution (definition)


Why use simple models for Machine Learning?

Linear regression versus Logistic Regression differences and similarities.

Machine learning classes offered at Stanford.


Our writing on medium Famous Machine Learning Datasets You Need to Know [medium] include a blurb about the famous COCO dataset [definition]. What’s does an image classification dataset look like? Almost all data needs massaging before being fed into neural networks / models. Here are some tips on how to handle missing data. Pop quiz : Question / Quiz: Is dimensionality reduction data preprocessing, Supervised Learning, Unsupervised Learning Or reinforcement learning?

Pytorch basics: explanation of Pytorch sequential container. Intermediate Pytorch tooling: Pytorch Profiler Performance Tool. Here are some [pro] tips on how to get started with torch profiler.

Low-code Machine Learning in Google Cloud: Did you know you can use BigQuery for Machine Learning? Just write some SQL — a query language that most data analysts and data scientists already know. BigQuery ML (BQML) Workflow. More about BQML BigQuery Machine Learning

Machine Learning and Deep Learning in the Cloud

AWS SageMaker workflow and capabilities. Why use Amazon SageMaker? Cloud GPUs cloud computing for AI projects? Use cases. If you do decide to go with AWS, here are some pro tips for choosing compute on AWS [pro]. It’s easy to deploy XGBoost on AWS [pro]. Once developers deploy models, they can monitor the models using AWS SageMaker Debugger [pro]. (Reminder disclaimer: none of our content has been tested in production. Our content is not for live/deployment purpose. Informational educational purpose only. )

Solving Challenging Machine Learning Problems at Tesla

Lane detection, camera challenges for Tesla.

Tesla AI Day

Natural Language Processing

In today’s NLP, transformers actually perform many NLP tasks in companies. (Disclaimer : read our full disclaimer, none of our tutorials are production ready. They are for personal, informational purpose only.) What is a transformer? Transformer Basics. Best explanation for Transformer [pro, easter egg, definition]. Best explanation for the previous state-of-art RNN [pro, easter egg, definition]. Recent transformer, BERT models you should know [pro]. More transformer models developers should know [pro]. Working with SpaCy? Here’s how to create a custom component. Working with Hugging Face? The current state of art models? Check out this [pro] flash card Hugging Face state of art NLP models — beginner tutorial with pro tips for developers. Check out this easter egg NLP state-of-art transformers tutorial.

A review of classic NLP techniques : n-gram definition, n-gram definition 02


A review on GAN . How important is prompt design when using GPT-3 and other generative models? What are the best practices? We will cover this in the next newsletter.

Stable Diffusion is the power engine behind the recently trending ultra high definition, hyper realistic generated arts. Some of these digital arts almost look fantastical. There’s suddenly an influx of Stable Diffusion tutorials. Here’s what we recommend as the high quality tutorial on Stable Diffusion [easter egg]. What is stable diffusion? [pro] What’s stable diffusion II (definition)? [pro]

Here’s a funny meme about Generative Adversarial Networks.

Google Imagen


Web Development Skills, Developer Tools

React Basics : React Router navigates the apps and the pages. A commonly used task and tool for React developers is bundling [definition], this helps combine assets and improve app performance. There’s also React Redux.

Other dev tools:

Subscribe now


There are classical data structure and algorithms interview questions. There are also interview questions about newer and trendier technologies: such as graph database and graph algorithms, which took off with social media companies like Facebook and are omnipresent.

How to make a developer laugh? Here are some developer jokes and memes.

Big O Developer Jokes.

Funny meme about GAN Generative adversarial network

What is the hot dog not hot dog app?

Here’s a freebie for developers.

Professor Andrew Ng’s free ebook : How to build a career in AI book

Greedy algorithms may not always arrive at the optimal solution, but they belong to an important category of algorithms, and often comes up during technical interviews (for example, it may be important to spot that the greedy algorithm is not optimal).

Career — Layoffs

We wrote extensively about handling rejection and layoffs in tech in previous newsletter. Disclaimer: we are not career counselors this is not professional advice. Not all companies require computer science degrees. For example during the Lesbian Who Tech conference, job opportunities questionnaires ask for certificates instead of degrees. (By the way, the conference also offered child care center for attendees. We see that as very helpful for visiting mom engineers. Shows progress in tech inclusion). Finally some certificates are useful! They don’t always give the best ROI. As a bootcamp graduate myself, I believe in the power of certificates, but I think the society and Silicon Valley has a long way to go to accept them. Recently there have been positive changes. But recent tech layoffs surely didn’t help.

Making sense of layoff according to a Stanford professor .

Could this be Google’s career rating ranking for developers


Tesla Engineering, Tesla AI Challenges — Lane Detection

Tesla AI Challenge Overview
Tesla Engineering, Tesla AI Challenges

Video Example of Why Self Driving is a Hard Problem to Solve

Merry Christmas and Happy New Year!!

Appendix links to new flash cards:

R Squared (definition)

Graphical Representation of R squared

Dev tools that data scientists use

Most popular languages among data scientists

Data science book resources [pro]

Correlation [definition]

Best explanation definition of RNN [easter egg, definition, pro]

Transformer Basics [definition]

Best explanation definition of RNN [easter egg]

Pytorch sequential container

Pytorch Profiler Profiling Tool Performance Tool

Using PYTORCH PROFILER (code snippets)

BigQuery ML (BQML) Workflow

BQML BigQuery Machine Learning

L2 Norm ||u|| calculating norm, distance [public, linear algebra, dot product]

What is normalization normalizing (definition) SQL, relational database

What is normalizing? (Database, SQL)

Famous Machine Learning Datasets You Need to Know [medium]

COCO Dataset — Machine Learning Dataset (definition)

Easter egg : MongoDB Pro Tip [pro]

Geocoding, geo data, Google Geocoding API (definition)

Edge device battery [pro]

Google Cloud, Cloud Functions [definition]

React Basics

React Router (definition) [definition, use case, pro]

React Redux

What’s Bundling (definition)?

Machine Learning Classes at Stanford

The Visual Studio Code WSL extension [public] Developer Tool. The Visual Studio Code WSL extension [for windows users]

Missing data best practice

Question / Quiz: Is dimensionality reduction data preprocessing, Supervised Learning, Unsupervised Learning Or reinforcement learning?

Graph Algorithms Basics in Technical Interviews

AWS SageMaker workflow, AWS SageMaker capabilities

Pick AWS SageMaker compute [pro]

Amazon SageMaker + XGBoost [pro, definition]

Why use Amazon SageMaker? Cloud GPUs cloud computing for AI projects? Use cases.

Create SpaCy NLP custom components

What is stable diffusion (definition) ?

Heatmap [pro]

Data visualization for confusion matrix

AI at Tesla [pro]

Funny meme about GAN Generative adversarial network

Linear Regression vs Logistic Regression

Why are simple machine learning models still in use?

Pro tip study tips from Andrew Ng Founder of Coursera [pro]

Comic about living with ADHD

Important models at OpenAI [pro]

Transformer Models, BERT related models [pro]

Transformer models you should know [pro]

Generative adversarial networks (GANs) [definition, basics]

Getting started with Hugging Face state of art NLP models, transformers [pro]

How does TikTok do machine learning?

Pro tip: understanding confusion matrix, precision, recall, sensitivity, specificity.

Make the most out of Kaggle data competition site [pro]

What’s does an image classification dataset look like?

Getting started with R Programming, RStudio

Greedy algorithm (definition)

AWS Sagemaker debugger, train, tune, monitor

Question / Quiz: compute on google

n-grams definition

what is n-gram ngram?

Easter egg: NLP state-of-art transformers tutorial

Benefits of using Google Cloud Platform as the Machine Learning Platform

What’s stable diffusion (definition) II [pro]

Google Cloud compute options

Uniqtech Data Science Bootcamp is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.