Similarity and competence measures to dynamic predictor selection:: a comparative study

Project description

This project evaluates the impact of alternative measures on dynamic predictor selection performance to predict one-step-ahead microservices time series.

Installation

How to install the project?

$ virtualenv venv
$ source venv/bin/activate
$ pip3 install -r requirements.txt
$ apt install zip -y; unzip pickle.zip;

Project Files

Summary of the main repository files.

Files	Content description
series-descriptions.csv	Description of the datasets.
result	MSE restults.
pickle	Trained models saved in .pickle

How to regenerate results using pickle models?

File	File description
competences.py	Competence measures.
dynamic_selection.py	Run the DS dynamic selection algorithm.
results.py	Generates csv results from models.
generate_oracle.py	Generate results from Oracle.
similarities.py	Similarity measures.
train_models.py	Train monolithic models and the homogeneous pool.