Train deep learning model for univariate anomaly detection
Budget: €30 – €250 EUR
Labeling of the already generated univariate dataset (based on types of anomalies)
Train deep learning model (except CNN) for univariate anomaly detection. (Preferred hybrid methods like autoencoder, lstm-ad, lstm-vae, mscred)
Protocol the evaluation metrics such as f1-score and computational time.
Also evaluate the precision, f1 score and computational time of the database with an existing repositories (https://github.com/HPI-Information-Systems/TimeEval-algorithms) containing the different neural network algorithms
Train deep learning model (except CNN) for univariate anomaly detection. (Preferred hybrid methods like autoencoder, lstm-ad, lstm-vae, mscred)
Protocol the evaluation metrics such as f1-score and computational time.
Also evaluate the precision, f1 score and computational time of the database with an existing repositories (https://github.com/HPI-Information-Systems/TimeEval-algorithms) containing the different neural network algorithms