A simple MLP neural network to predict PM2.5 and PM10 in Thailand. The model uses the day of the week and the month of the year as the main features. The model is not optimized and can be optimized by migrating from scikit-learn to Keras and MLP to RNN model.
Data extraction from the COVID-19 global data. This model combined unsupervised clustering and dimensional reduction technique. The model compares each clustering technique: K-Means, DBSCAN, MeanShift, and SOM model.