The Data Science skill no one talks about...

Every aspiring data scientist I talk to thinks their job starts when someone else gives them:
1. a dataset, and
2. a clearly defined metric to optimize for, e.g. accuracy

But it doesn’t.

It starts with a business problem you need to understand, frame, and solve. This is the key data science skill that separates senior from junior professionals.

Let’s go through an example.

Example

Imagine you are a data scientist at Uber. And your product lead tells you:

👩‍💼: “We want to decrease user churn by 5% this quarter”

We say that a user churns when she decides to stop using Uber.

But why?

There are different reasons why a user would stop using Uber. For example:
1. “Lyft is offering better prices for that geo” (pricing problem)
2. “Car waiting times are too long” (supply problem)
3. “The Android version of the app is very slow” (client-app performance problem)

You build this list ↑ by asking the right questions to the rest of the team. You need to understand the user’s experience using the app, from HER point of view.

Typically there is no single reason behind churn, but a combination of a few of these. The question is: which one should you focus on?

This is when you pull out your great data science skills and EXPLORE THE DATA 🔎.

You explore the data to understand how plausible each of the above explanations is. The output from this analysis is a single hypothesis you should consider further. Depending on the hypothesis, you will solve the data science problem differently.

For example…

Scenario 1: “Lyft Is Offering Better Prices” (Pricing Problem)
One solution would be to detect/predict the segment of users who are likely to churn (possibly using an ML Model) and send personalized discounts via push notifications. To test your solution works, you will need to run an A/B test, so you will split a percentage of Uber users into 2 groups:

The A group. No user in this group will receive any discount.

The B group. Users from this group that the model thinks are likely to churn, will receive a price discount in their next trip.

You could add more groups (e.g. C, D, E…) to test different pricing points.

In a nutshell

1. Translating business problems into data science problems is the key data science skill that separates a senior from a junior data scientist.
2. Ask the right questions, list possible solutions, and explore the data to narrow down the list to one.
3. Solve this one data science problem

Wanna learn more real-world Machine Learning? Join @Artificial_intelligence_in

Data scientists spend 80% of their time working on the data.

Books spend 80% of their time talking about algorithms.

Today, there's a large gap between academia and reality. Between what they say is important, and what really is.

Better data is better than better models.

Ai advancement in 2023

January: AI progress in medicine & healthcare; OpenAI's & Microsoft deal

February: ChatGPT's success; Google's Bard A.I.; Microsoft's new Bing with ChatGPT

March: Adobe's Firefly; Canva's AI tools; OpenAI's GPT-4 & new APIs

April: ChatGPT with Boston Dynamics' robots; My AI on Snapchat

May: Advances in humanoid robots; FDA approval for Neuralink; creative AI developments

June: Apple's Vision Pro AR headset; OpenAI's non-U.S. office plans; AI in drug discovery

August: ChatGPT's 'Custom Instructions'; Google's AI Genesis; AI copyright discussions

September: OpenAI's web browsing in ChatGPT; Midjourney & Stability AI's creative tools; AI enhancements in YouTube & Amazon

October: ElevenLabs' AI translation technology;Dawn Phase 1

November: xAI's Grok launch; OpenAI's Assistants API, GPT-4 Turbo; Meta's AI division changes; AI Safety Summit

December: Google's Gemini release; AMD's AI hardware advancements; EU's AI Act

No doubt about 2023 was the year of Artificial Intelligence.

Dear Community,

As the year draws to a close, I would like to express gratitude for being a part of the Ai India community dedicated to expanding AI education.

My prediction for 2024💡

1. 1B models will outperform 70B models.

2. Models will be deployed on CPUs for almost free. Not API services.

3. Data quality will yield the next 10x boost in performance.

4. A combination of open source models will beat the best private models.

5. Compilers will unlock at least 80% speed up in models (both training and inference)

6. Legislation will side with content creators over model developers.

7. Local ML is going to be huge. It will be in part driven by the adoption of Apple Silicon and other innovative hardware, but also on raw CPU and mobile devices

8. In many cases except for the largest of LLMs, local inference will become a viable alternative to hosted inference.

Join WhatsApp Channel 🚀

Happy New Year 2024 🎊🎉

Here are 300 hours of curated courses focused on Machine Learning Engineering.

There are 15 courses.
From beginner to advanced.
From Google.
For free.

Some of the topics they cover:

• Fundamentals of Machine Learning
• Feature Engineering
• Production Machine Learning Systems
• Computer Vision and Natural Language
• Recommendation Systems
• MLOps
• TensorFlow, Google Cloud, VertexAI

The courses are well structured. They aren't just links to YouTube videos. You have to join the course, and they have an interface that takes you through every module.

This is good content. And it's free. https://www.cloudskillsboost.google/paths/17

I teach hard-core Machine Learning Engineering. 🚀

Artificial Intelligence changed my life forever. There has never been a better time to build a career that will set you apart for the next 20-30 years.

I teach a program where I show people how to build Machine Learning systems.

My program is not an online course. It's not a group of videos you watch and a PDF you read.

My program is a 90-hour live class with an additional 13 Assignments and 9 projects material. It's tough.

While everyone wants to know what will happen in the next ten years, we won't waste time trying to predict that. Instead, we focus on what never changes.

The program is about the fundamental principles of building machine learning systems. It's about timeless ideas that will help you understand the future, whatever that is.

My guarantee is simple: you'll learn more than you've ever done before.

• Cohort #1 starts this Saturday on 13th January

You can join here: www.Aiindia.ai/bootcamp.

Some of the most frequently asked questions:

How much do I have to pay?

The cost to join is a one-time payment of ₹30k and for students ₹10k There are no recurrent payments. Once you join, you get lifetime access to all materials and a community of engineers who went through the program.

Is every class live, or can I watch them offline?

Classes are live, but you can watch the recordings at your own pace.

What are the prerequisites to join?

Ideally, you don't need any prerequisite to join our program only required is your dedication and commitment. You don't need a machine learning experience to learn.

What are some of the topics you'll cover?

We cover a lot, but here are 10 of the most important topics we'll discuss in class:

1. Framing machine learning problems
2. How to fine tune models and transfer learning.
3. Processing, training, deploying, inference pipelines
4. Offline evaluation and testing in production
5. Performing error analysis. Where to work next
6. Distributed training. Data and model parallelism
7. Pruning, quantization, and knowledge distillation
8. Model deployment. Online and batch inference
9. LLMs basics with training.
10. Python programming from scratch.

Here is the link to join: www.aiindia.ai/bootcamp

Forget Dally2 and Use HidingElephant🔥

Use one of World's Best AI logo Designer Tool, HidingElephant can Quickly turn text prompts into logos, easily convert designs to vectors, and generate multiple concepts from one idea. It's ideal for designers seeking efficient, creative solutions for contests, client work, and team projects.

Check it out at: https://ai.hidingelephant.com/AiCommunity

Sign up to enter a draw for a free ChatGPT-4 premium subscription.

AI startups raised ~$50B in 2023 alone. Just WOW.

This is a class from Harvard University:

"Introduction to Data Science with Python."

You can take this class for free. If you want the certificate, you can pay $299 for it. You should be familiar with Python to take this course.

The course is for beginners. It's for those who want to build a fundamental understanding of machine learning and artificial intelligence.

The course will cover some of these topics:
• Generalization and overfitting
• Model building, regularization, and evaluation
• Linear and logistic regression models
• k-Nearest Neighbor
• Scikit-Learn, NumPy, Pandas, and Matplotlib

This course is perfect if you are a software developer with experience writing Python code and want to start with Machine Learning.
Link: https://pll.harvard.edu/course/introduction-data-science-python

Excited to share with you some FREE but very useful NEW AI Courses!

These are great for anyone who wants to dig deeper into the topic this year.

𝟭. Prompt Engineering Basics: https://explore.skillbuilder.aws/learn/course/external/view/elearning/17763/foundations-of-prompt-engineering

𝟮. ChatGPT Prompts Mastery: https://www.deeplearning.ai/short-courses/chatgpt-prompt-engineering-for-developers/

𝟯. Intro to Generative AI: https://www.cloudskillsboost.google/course_templates/536

𝟰. AI Introduction by Harvard: https://pll.harvard.edu/course/cs50s-introduction-artificial-intelligence-python/2023-05

𝟱. Microsoft GenAI Basics: https://www.linkedin.com/learning/what-is-generative-ai/generative-ai-is-a-tool-in-service-of-humanity

𝟲. Prompt Engineering Pro: https://learnprompting.org

𝟳. Google's Ethical AI: https://www.cloudskillsboost.google/course_templates/554

𝟴. Harvard Machine Learning: https://pll.harvard.edu/course/data-science-machine-learning

𝟵. LangChain App Developer: https://www.deeplearning.ai/short-courses/langchain-for-llm-application-development/

𝟭𝟬. Bing Chat Applications: https://www.linkedin.com/learning/streamlining-your-work-with-copilot-formerly-bing-chat-bing-chat-enterprise/put-your-fingers-to-work-chatting-as-a-productivity-tool

𝟭𝟭. Generative AI by Microsoft: https://learn.microsoft.com/en-us/training/paths/introduction-generative-ai/

𝟭𝟮. Amazon's AI Strategy: https://explore.skillbuilder.aws/learn/public/learning_plan/view/1909/generative-ai-learning-plan-for-decision-makers

𝟭𝟯. GenAI for Everyone: https://www.deeplearning.ai/courses/generative-ai-for-everyone/

𝟭𝟰. AWS GenAI Foundation: https://www.coursera.org/learn/generative-ai-with-llms

♻️ 𝗕𝗼𝗻𝘂𝘀:

• OpenCV Bootcamp: https://opencv.org/university/free-opencv-course/

• Tensorflow Bootcamp: https://opencv.org/university/free-tensorflow-keras-course/

StackOverflow is a dinosaur that's going extinct.

I mostly use Copilot, ChatGPT, & Perplexity. I haven't found a single reason to visit StackOverflow anymore.

StackOverflow can't compete, & I don't see how they can stay relevant any longer.

I use Copilot for inline suggestions. As I type, Copilot takes care of the little things. It saves me from dozens of searches every day.

ChatGPT is the workhorse. I use it to solve more complex tasks with my code. Here are some examples:

• Explain what this code does
• Simplify it
• Rewrite it in a more efficient way
• Rewrite it in a more readable way
• Replace the use of a library with another
• Write documentation for it
• Describe potential edge cases
• Write unit tests for those edge cases

I use Perplexity to ask questions. Google is broken. If you don't believe me, try Perplexity for a day.

(Google is still king as a navigation tool, I never type a complete URL in my browser)

Modern AI-powered tools are replacing boomer tech.

StackOverflow is dead to me.

Now Bard becomes Gemini 🚀

Google launches Gemini Ultra, its most powerful LLM yet

👉 Google renames Bard to Gemini, launches paid ‘Gemini Advanced’ with Ultra 1.0

👉 Gemini is the company’s much-anticipated response to ChatGPT and OpenAI’s GPT-4 model, which was the industry leader to this point.

👉 Based on my early testing, Gemini Ultra matches or exceeds GPT-4’s performance. In particular, there’s one thing that Gemini Ultra does way better than ChatGPT.

👉 Gemini’s release is Google’s biggest AI move in at least a year, and Gemini is now ChatGPT’s only true competitor.

👉 On their official launch page, Google provides three examples of the new things Gemini Ultra can do. I decided to test the model’s performance by trying each of those thing and found that results are really good and promising.

⭐️ I have canceled my OpenAI Premium subscription and now using Gemini's 2-month FREE trial—YES, a free trial is indeed available for two months.

Checkout: https://gemini.google.com/app

To Know more: https://blog.google/products/gemini/bard-gemini-advanced-app/

⭐️ A simple guide for those who want to learn machine learning:

These are 3 online courses you can take in order. I also have a bonus for those who need a little bit more.

First, start with freeCodeCamp's Python course.

It's a free 9-hour course. You'll find it on their YouTube channel. Search for "Python Tutorial for Beginners (with mini-projects)."

Second, cover the basics of machine learning.

Take Google's Machine Learning Crash Course. It's free, and you'll find it online.

This course starts from scratch, and it's focused on beginners. It's an excellent choice for those who want to take things slow.

There are two alternatives that I recommend to people who want to take it even slower:

• Intro to Machine Learning
• Intermediate Machine Learning

These are tutorials from Kaggle. These are short courses, but you will learn a ton from them. 

Finally, take the Machine Learning Specialization on Coursera.

Completing this specialization will take several weeks. It's not free, but nothing that matters is.

Bonus for those who need a bigger challenge:

Explore GitHub projects, read top research papers in the field. Remember real world learning is always important.

In summary: 🚀
1. Python Tutorial for Beginners (freeCodeCamp)
2. Google's Machine Learning Crash Course
3. Coursera's Machine Learning Specialization

But what about the math? 📊

Don't rush it. You'll have plenty of opportunities to learn mathematics. Start building and pick the math as you go.

People always ask me about the timeline: How long will this take them?

My answer is always the same: how long do you have? This will take all of it.

Happy Machine Learning! 🤖

This week’s AI news:

🎥 All about OpenAI’s Sora.

🤟Yolov9 is out

🖥️ Huawei's AI chips gain traction.

🤖 Google Gemini 1.5 is out

🤸 AI judges gymnastics with precision.

📸. Google Gemma - New Open-Source LLM #ainews

I plan to share AI news updates every saturday. If you support this AI news initiative, Show your support 👍

This week's Ai News

🧠 Krutim AI's chatbot demo failed, It said it's built upon OpenAi

🚗 Apple Car Cancelled

📰 Google Quietly Paying Journalists to Generate Articles Using Unreleased AI

👊 Elon Musk Sues OpenAI

🗞Mistral CEO confirms ‘leak’ of new open source AI model nearing GPT-4 performance.