Everything I Studied to Become a Machine Learning Scientist at Amazon from ZERO Tech Background

Summary

In this video, the speaker shares their seven-year journey from a political science graduate with no technical skills to an applied scientist at Amazon specializing in machine learning. They detail the resources, courses, and books that helped them learn essential skills such as Python, machine learning fundamentals, and system design. Emphasizing the importance of learning how to learn and applying skills through projects, the speaker offers advice for non-technical individuals aiming to break into tech, sharing both high ROI investments and time-wasting lessons learned along the way.

Takeaways

• 😀 Your background doesn't define your future in tech – with intentional learning and consistent effort, anyone can transition into technical roles.
• 😀 Start with the basics – Python is essential, and focusing on it early in your learning journey is crucial for building a strong foundation in tech.
• 😀 Project-based learning is key – applying your skills through personal projects accelerates your progress more than any single course.
• 😀 Strong math skills, especially in calculus, linear algebra, and probability theory, are essential for deep understanding of machine learning.
• 😀 Learning to learn efficiently is more valuable than any specific technical skill – building strong study habits will help you adapt to new concepts throughout your career.
• 😀 The importance of interdisciplinary knowledge – being able to bridge technical concepts with business or social problems is an advantage in the tech field.
• 😀 It's important to understand the bigger picture – system design, MLOps, and scalability are crucial as you advance in your career.
• 😀 Certifications and courses are helpful, but practical, hands-on learning (through projects and real-world applications) is the most effective way to internalize skills.
• 😀 Some resources, like Hadoop or certain data visualization tools, have a low ROI for practical work and may be better learned on the job.
• 😀 Non-technical fields can provide transferable skills that make you more versatile in tech roles, such as the ability to translate complex ideas into business solutions.

Q & A

What was the speaker's background before entering the field of machine learning?

The speaker had a background in political science and public policy, with no technical skills. They were working at a jewelry company before transitioning into machine learning.

How did the speaker first get introduced to programming and machine learning?

The speaker first got introduced to programming by purchasing courses on Udemy, including 'The Complete Python Bootcamp' and 'Python for Data Science and Machine Learning.' They started learning Python and machine learning basics during their master's program in public policy.

What courses did the speaker take during grad school to build technical skills?

During grad school, the speaker took courses in traditional statistics with R, causal inference, Python programming, machine learning fundamentals, deep learning, and NLP. These courses were part of their social science curriculum and were technically oriented.

What was the speaker's thesis project about, and how did it relate to machine learning?

The speaker's thesis project involved predicting who someone voted for based on their web browsing history. This was an application of machine learning techniques for predictive analysis.

How did the speaker supplement their learning outside of formal education?

The speaker supplemented their learning with self-study, including online courses on Tableau, PowerBI, SQL, data science interview prep, Hadoop, and statistics. They also watched a statistical rethinking course on YouTube.

What did the speaker learn from their data science internship and subsequent job at Coursera?

During their data science internship at Coursera and their subsequent job there, the speaker gained access to unlimited free courses. They revisited machine learning fundamentals and took various courses to deepen their knowledge, such as 'Learning How to Learn,' Stanford's machine learning specialization, and deep learning courses.

What role did math play in the speaker's journey, and how did they improve their skills?

Math, particularly linear algebra, calculus, and statistics, played a crucial role in understanding machine learning algorithms at a deep level. The speaker improved their math skills by taking the 'Math for ML' specialization and using resources like 'Three Blue One Brown' and 'StatQuest.'

How did the speaker's time at Twitch and Amazon influence their learning?

At Twitch, which is owned by Amazon, the speaker had access to AWS training and certifications, which they used to deepen their understanding of machine learning. They also studied system design, which became a focus of their learning in recent years.

What technologies and concepts did the speaker focus on during their studies of machine learning system design?

The speaker studied Java, data structures, algorithms, and system design to improve their understanding of machine learning system design. They took various courses, including those from Udemy, Coursera, and Georgia Tech, and worked through books and courses on system design and machine learning systems.

What were the speaker's thoughts on the AWS machine learning certification, and how did it impact their career?

The speaker felt that the AWS machine learning certification, although valuable in some ways, did not significantly improve their practical knowledge. They found the process of preparing for it to be stressful and focused too much on rote memorization. They believe real project-based learning would have been more beneficial.