About this Course
Within the first course of Machine Studying Engineering for Manufacturing Specialization, you’ll establish the assorted parts and design an ML manufacturing system end-to-end: venture scoping, information wants, modeling methods, and deployment constraints and necessities; and discover ways to set up a mannequin baseline, tackle idea drift, and prototype the method for growing, deploying, and constantly bettering a productionized ML utility.
WHAT YOU WILL LEARN
Determine the important thing parts of the ML lifecycle and pipeline and evaluate the ML modeling iterative cycle with the ML product deployment cycle.
Perceive how efficiency on a small set of disproportionately vital examples could also be extra essential than efficiency on the vast majority of examples.
Resolve issues for structured, unstructured, small, and massive information. Perceive why label consistency is important and how one can enhance it.
SKILLS YOU WILL GAIN
- Human-level Efficiency (HLP)
- Idea Drift
- Mannequin baseline
- Mission Scoping and Design
- ML Deployment Challenges
Syllabus – What you’ll be taught from this course
5 hours to finish
Week 1: Overview of the ML Lifecycle and Deployment
This week covers a fast introduction to machine studying manufacturing programs specializing in their necessities and challenges. Subsequent, the week focuses on deploying manufacturing programs and what’s wanted to take action robustly whereas dealing with consistently altering information.
3 hours to finish
Week 2: Choose and Prepare a Mannequin
This week is about mannequin methods and key challenges in mannequin improvement. It covers error evaluation and methods to work with totally different information sorts. It additionally addresses how to deal with class imbalance and extremely skewed information units.
4 hours to finish
Week 3: Information Definition and Baseline
This week is all about working with totally different information sorts and guaranteeing label consistency for classification issues. This results in establishing a efficiency baseline on your mannequin and discussing methods to enhance it given your time and assets constraints.
0 Comments