.jpeg)
As we always say, "practice makes the master". It's not enough to "read" or watch videos about Machine Learning. At some point, it's important to start writing some code to apply that knowledge.
In this article, we introduce 7 free Machine Learning projects that will help you practice your skills in an incremental way, from the most basics, to the most advanced. Let's get started!
Let's start with a broad overview of the process that we conduct when starting a Machine Learning project. You might think that it's all about creating deep neural networks and optimizing parameters. But everything starts a step before that. For example, the process of Data Cleaning for Machine Learning can take UP to 60% of the total time of the whole project.
Our first recommended project is: Predicting intergalactic transportations with Spaceship Titanic. It's a guided project that takes you all the way from importing the data, analyzing it, cleaning it, rearranging it, until it's ready to train a Machine Learning model on it. This project is a GREAT example of the lifecycle of Machine Learning in real life.
If you want to get started with something shorter and to the point, we recommend Can you classify the monsters that are haunting?. This project is short and to the point: identify the target variable, do some basic analysis, and finally train a classifier. Similarly you can find Happiness Classification interesting. This project makes special emphasis in splitting and preparing the data, but then it goes right to the point and tasks you with training a LogisticRegression classifier.
Some basic Regression projects include: Multiple linear regression: CO2 emissions in vehicles and Predicting Fuel efficiency of old cars.
Once you have finished analyzing, cleaning and re-arranging your data, it's time to train the Machine Learning models. But sometimes, you'll find that models out of the box are not as effective as you thought they'd be. That's why we apply "hyperparameter tuning". Which basically tries combinations of different parameters of the model in order to get the best results for our use case. A great project to try this, is Hyperparamter Optimization when classifying monsters. In this project you'll apply GridSearchCV and RandomizedSearchCV to find the best parameters for a Decision Tree.
If you're up for a tougher challenge, you could try our Customer Churn Prediction using XGBoost project. It is a very complete (and sorry, a bit long) project that takes you through all the process from analyzing and cleaning the data, up to tuning the parameters of an XGBoost model.
That's it! Leave us a ranking and tell us what you think about these projects!
