You NEED to learn Machine Learning, RIGHT NOW!!

About this video

The creator argues machine learning can be learned without a degree or advanced math, using advice from machine learning engineers Santiago, who works on computer vision for Boston Dynamics’ Spot robot according to the video, and Nacho, who built the hands-on labs. Machine learning is defined as teaching a computer to learn rules from labelled data instead of hand-coding rules, illustrated by training a convolutional neural network on photos labelled as the creator and not him. The path is learn-by-doing: sign up for a free Oracle Cloud Infrastructure account — the video says signup offers a $300 credit — and work through Nacho’s three workshops, described as about one and a half to two hours, building a League of Legends win predictor. Learners deploy infrastructure with Terraform and Ansible, pull match data from the Riot Games API with a free key, and work in OCI Data Science Jupyter notebooks using tools named in the video such as scikit-learn, AutoGluon, pandas, NumPy and Matplotlib. The creator stresses data extraction and preparation is, per Santiago and Nacho, 70 to 80 percent of the job. Further steps are intermediate Python, Kaggle’s free Intro to Machine Learning and Kaggle competitions, brushing up high-school-level statistics, probability and calculus via Brilliant or Khan Academy, and Andrew Ng’s Machine Learning Specialization on Coursera.

Key takeaways

• A model is a file trained on labelled data with a chosen algorithm; the engineer’s job is improving data, features and algorithm choice when accuracy is poor.
• Start by doing: the OCI League of Legends labs move from Riot API data extraction to training a win-prediction model in Jupyter notebooks.
• Data extraction and preparation dominates the work — the video attributes a 70 to 80 percent share to Santiago and Nacho.
• Intermediate Python scripting is required for pulling API data and shaping it, but not full software-engineering expertise.
• Follow-up path named in the video: Kaggle’s Intro to Machine Learning, high-school-level math refreshers, and Andrew Ng’s Coursera specialization, then portfolio projects.

Episode details: published March 1, 2023 · 16:46 long. Summary written by Self Growth Videos from the episode itself. Business results and any income figures mentioned are NetworkChuck’s own claims, shared for education — they are not typical or guaranteed.

← Previous: why are TUPLES even a thing? · Next: the most SECURE browser!! (testing it with malware) →

All NetworkChuck videos · NetworkChuck profile · NetworkChuck on YouTube