Assignements & in-class Kaggle competitions for the course Applied Machine Learning at NMBU.
A description of the different files/projects:
Electricity Price Prediction - 3.ipynb: The main task is to develop a ML model to classify whether electricity prices will increase or decrease in New South Wales and Victoria. The raw data is retrieved from the Australian Energy Regulator. The main goal is to use models from the sci-kit learn. This is a beginner project which only focuses on exploring, preprocessing, and creating a model with only using multiple train_test_splits while searching for the best hyperparameters instead of applying cross-validation techniques.
Predicting Real Estate Price Class - 4.ipynb: The main task is to develop a ML model to predict which price-class a property in the city Melbourne belongs to. The main goal of this classification project is to use the basic Ml techniques such as exploring, preprocessing (data cleaning, feature transformations), creating pipelines, hyperparameter- tuning, training, testing/validating and assessing the model based on ROC, accuracy, F1-score.
Rental Bike Usage Prediction - 5.ipynb: The main task is to predict how many bikes that will be used on any given day. The dataset consists of both numerical and categorical features. The target values are continuous and this is a regression problem. The main goal of this project is to use the basic Ml techniques such as exploring, visualizing, preprocessing (data cleaning, feature transformations), creating pipelines, hyperparameter- tuning, training, testing/validating and assessing the model based on R2. In addition the targets are transformed into bins in order to build a classification model on the same dataset used for the regression problem.