Machine Learning Newsletter March 2021 #2

Machine Learning & Deep Learning Tutorials on substack.

This newsletter will help you learn about machine learning much faster than googling on your own. Paid subscribers get the most out of these newsletter but there is plenty of free content floating around.

Read our disclaimer here : 👉🏻 disclaimer, policy, license.

Natural Language Processing is still the hottest trend

Trend : the rise of NLP continues

How many times is NLP mentioned in clubhouse? It is mentioned constantly in every chat room about AI and deep learning. And GPT-3 is the crowning jewel of that conversation.

Read about SpaCy (SpaCy basic view, SpaCy pro view), GPT-3 requires login but free. All our GPT-3 findings are compiled here, on this landing page. Our scholar also received GPT-3 access. What would you ask GPT-3 if you have access? In the next newsletter we will write about GPT-3 prompt design - a new way to interact with models using prompts - this is essentially the design paradigm of GPT-3. Message us on to ask your question for GPT-3.

What if you can’t access GPT-3, or are not admitted to GPT-3? Is there an alternative? We might have an answer to the question. Learn more here.  👉🏻 Next week, we will discuss a GPT-3 alternative.

What do you want to ask GPT-3?

Next week’s trend, what can you do get started with basic blockchain, smart contracts, non fungible tokens (NFT). We know every one has been talking about NFT, but we went to Stanford with the OGs who started dabbling with blockchain more than 5 years ago. We will explain in the next newsletter how as a data scientist, data engineer, machine learnist, you can get started with some mini tutorials to learn about the blockchain and crypto. Again read our disclaimer. All our contents are only for informational purpose only. They are not to be used in investing, nor as professional advice, nor in commercial settings.

Career Coaching - What does it take to become a machine learning engineer?

We are not career counselors but we collect and share insights from deep learning and machine learning experts. Here’s a quote from Mr. Laurence Moroney lead developer advocate at Google for Tensorflow, AI discussing how to get into Machine Learning, Edge Computing for Embedded Systems, TinyML in a recent Q&A online. He’s a truly great developer advocate and is great at communicating with developers. This detailed quote is for pro / paid members only. Visit the link here 👉🏻

The previous newsletter

Like what you read? Read the first part of the newsletter below. Note you will need to sign up for a free account on to access. Both newsletters are free, but only paid subscribers have access to premium skill cards (selected links in the tutorial)

To read Newsletter March 2021 #1 basic access visit here. 👉🏻 link to free March newsletter.

We have many more ML career skill cards publishing soon.

To read Newsletter March 2021 #2 pro access visit here. 👉🏻 link to pro march newsletter.

Already a pro member? Message us in the app to report any issues.

Developer Lifestyle

Why are people so fascinated with FAANG (Facebook Apple Airbnb Netflix and Google) companies? It might be because of the amazing perks these company employees would enjoy. Developers are treated well at conferences, seminars, recruiting events before the pandemic hit. Here are some highlights, company “virtual tours” and notable gifts and swags. Virtual company tours coming this week.

Github (coming soon)

Airbnb (coming soon)

Netflix (coming soon)

Apple (coming soon)

Facebook (coming soon)

Google (coming soon)

Developers get the best swags, gifts from companies. Here’s what we got. Do you have any amazing swags you’d like us to feature? Facebook hackathon swags. We will share a story about a data bootcamp in paradise (pre COVID). Too broke to buy masks? Major League Hacking gives out free masks and developer stickers.

Developer lifestyle can be a lot of hard work, but it can also mean being a well paid digital nomad, travel with code. We have a lot to share about our lucky travels with code and tech conferences in the next newsletter.

Is it too late to join San Francisco and Silicon Valley? Tech Exodus explained and monitored. #trend This card is for pro, paid subscribers only.

End-to-End Machine Learning Course

Our end-to-end machine learning course content will be released soon. There will also be a technical interview course using python. For pro members. Some free contents are available. Stay tuned. Pro members look for a separate newsletter about this week’s Machine Learning Course content. Multiple mini tutorials will be shared with you this week.


Need a review for Python list comprehension? Read it here. Basic access available to all visitors.

Install SpaCy and get started with NLP. Also basic access.

Get the most out of conferences . Coming soon get the most out of hackahtons.

Stay in Touch

Contact Us : email us or message us in the app (preferred) Like what you read? Buy us coffee Coffee counts as one month subscription.  Follow us on twitter. Follow us on Medium

OpenAI, Elon Musk’s GPT-3 — hottest new model explained!

Frontier of Natural Language Processing

[NEW] Our take on #OpenAI GPT-3

GPT-3 Past, Present and Future of AI and NLP 

link here 👉🏿 👉🏽 👉🏻 

Thanks for being a subscriber. Have you heard of OpenAI’s GPT-3? Simply put, it may be the most exciting tool for Natural Language Processing (NLP) in 2020. It’s the hottest and trendiest. In this NEW article, we give an overview of the model, use case, and metadata. More details on the paper coming soon! Hope it saves you time.

Don’t have a Medium premium account? We are making this article free for all subscribers. Request your link here

Like what you read? Buy us a cup of ☕️

#GPT3 #DeepLearning #AI #Tesla #ElonMusk #NLP

Convert data into tensors - End to End Machine Learning Course 11

Both structured and unstructured data may need help converting to numbers. Structured data include texts, strings. Unstructured data can be documents, files. Machine learning models consume numeric data as input, and specifically the data is efficiently loaded as Tensors, parallel processed if applicable, and Tensor objects usually come with auto gradient capabilities. Tensor is a data structure that modern deep learning libraries use. Each may be implemented or used a bit differently. It is usually a multi-dimensional matrix with functionalities that is beyond math and is useful for deep learning compute and data engineering such as auto gradient compute. Updated August 2020.

Where does this course fit in the machine learning workflow? The purple dot indicate the position. In addition Gather & Clean Data, we also want to encode or convert data in this step.

Encoding is a large field and with much complexity (It can be used to make information more compact, encode parameters into URI URL etc.). Right now we are just focusing on converting frequently seen data types into numbers they can be consumed by popular machine learning models. Pro tip : always check the dimension, sequence (order of inputs), and type of inputs the model architecture is expecting. These numbers don’t lie and they can give you a hint to how to fit something complex together. Side note: encoding is not to be confused with auto encoder a model architecture.

Our algorithm is numeric but the input data won’t always be.

Converting text, categorical and other complex data into numeric values is a true challenge. And continues to be a research area.

Categorical encoder / categorical data encoder

One hot encoding

Prefer having the code to try this out? Paid subscribers request your code here.

The output dimension of one hot encoder is row number - the number of data samples, column number - the number of unique values. In natural language processing (NLP), bag of words, it can be the number of unique words, often called the vocabulary.

One hot encoding assumes each of the label is independent from each other. For example, a sample cannot be both cat and dogs. A familiar example is that a coin is either head or tail (in none quantum computing world). It has to be either a cat or dog. There is no overlap of categories. It result in a sparse matrix, each row should have all entries as 0, except for the one corresponding, correct column label as 1. For example, if cat is the zeroth position [1,0,0, 0…..] there are are as many zeroes as there are unique values to encode. In the cat, dog, bird, horse example, the encoding for cat is [1,0,0,0], for dog is [0,1,0,0]. The dot product of two rows of different labels will always be zero. The dot product of two rows of the same label will always be 1.

"one hot encoding each row should only have 1 label while the others are 0"

Encoding data using one hot encoding can turn the original one column into many new columns, hence can take up space, time, cost and increasing search space, computing time, requiring more training data.

See paid subscriber only blog for One Hot Encoding in Pandas and an example of CountVectorizer which is a related concept in Natural Language Processing (NLP). Link here. Note you must accept the blogger invite to access the private blog.

It’s also important to use the above link to understand the difference between ordinal versus non-ordinal data. Basically one hot encoding is used for non-ordinal data.

Previously handling missing data and imputation strategy visuals also appear in the paid subscriber only blog. Link here.

ScikitLearn LabelEncoder

Prefer having the code to try this out? Paid subscribers request your code here.

Label encoder is a very simple encoder. It is much easier to understand than one hot encoding. It is also for categorical data. But it works better for ordinal data, or a pre-sorted input list, because its output imply that the labels are not strictly independent. Read more here, on the private blog. This can be confusing. We plan to have youtube videos to accompany these articles soon. Let us know if you want it sooner

Encoding Texts

In the private blog, we explain that one hot encoding can be used to encode texts. That is a basic yet useful technique. Modern day techniques are more advanced. It is called word embedding. Happy to explain more in the near future.

To understand more about embedding you can read our intro blurb on medium. Paid subscribers request free link here

Encoding Images

Turning images into tensors really warrants its own post so that will be explained in the near future as well. We will also cover scaling images, normalizing. To get a taste of encoding images as tensors read our intro to tensor post on Medium. Paid subscribers can request a copy for free. Understanding Tensors and Matrices.

Note on paid subscribers:

Paid subscribers get access to all our Medium and blog posts for the time of active paid subscriptions. Each paid period is one month. Private blog invitation is sent when the first payment is confirmed.

Scaling, Standardizing, Normalization

An important topic in the near future is scaling, which makes training data more stable, easier to compute, and easier to store. That’s why you see scikit learn utilities such as StandardScaler.

Loading more posts…