Ever confused on what Machine Learning (ML) is about, and what you should do as a student to learn more about Data Science? In this session, Traci shared more about ML and what can be done to learn about ML, through working on projects.
Feel free to take a look at some of the things that he has shared!
1. What is a ML Project
Some examples would include Creative Hackathon Quality, Best-Performing-Model Quality, Schoolwork Quality, Pet Projects and Professional Quality projects. Resources such as Github, Kaggle and Devpost have hackathons/projects that are shared and available online.
2. Process of a ML Project
There’s a series of steps in a ML Project.
Firstly, a problem definition needs to be crafted. It’s recommended that you can work with a team and discuss on a problem to solve (E.g. Image Recognition, Airline ticket analyser, Text Analyser, Music Recommendation). The problems doesn’t need to be unique, we’re just students learning from the process, and there’s a lot of resources available online that we can explore on!
You need to know what data you have. Data Cleaning may be required. Data Transformation may be required (change text into numbers). You can learn more about it on the go when you start working on projects!
How do you know how accurate is your algorithm? Matrixes would be required to calculate the accuracy levels of your project.
Features, also known as columns/dependent variables. They can be variables such as house type, prices, age, basically any input to the algorithm.
Have a formula to calculate and have the algorithm to include the formula and calculate the results.
Check for the results and see if they are what you would expect. If not, find out why, check if you would need to hyper-parameter tune our models!
Click here to learn more about what was gone through in the second half here!