Machine learning scripts and notebooks related to FIREMAN project.
Documentation: https://fireman-project.readthedocs.io
Artificial dataset used for development:
Repositories:
- Data imputation - data imputation scripts, notebooks and utilities
Scripts and notebooks:
- 00 Preprocessing - general approach to preprocessing of the dataset with exploratory data science(feature values/distributions/correlations/etc.)
- 01 Classification - basic application and result analysis of scikit-learn classifiers
- 02 Extended TEP - transformation of Extended TEP dataset in .rdata format to csv
- 03 Clustering - unsupervised density-based clustering analysis of the dataset using DBSCAN and OPTICS
- 04 Stream-based ML - MOA - stream-based machine learning using MOA
- 05 Pipeline - machine learning pipeline template
- 06 Stream-based ML - scikit-multiflow - stream-bades machine learning and concept drift detection using scikit-multiflow, project closely related to MOA
- 07 PCA - dataset analysis using PCA related to a collaborating student masters thesis
- 08 related analysis - dataset analysis related to a collaborating student masters thesis
- 09 PowerConverter dataset - Power Converter dataset preprocessing
- 10 PowerConverter dataset classification using DL - Power Converter dataset preprocessing
- 11 PowerConverter dataset classification using LightGBM - Power Converter dataset preprocessing
Additional materials:
- CSC/ - scripts usable in CSC
- workshop_05132020/ - materials and presentations related to Fireman ML workshop organized in May 2020
- adjusted Anomaly-Transformer scripts