Detects anomalies in time series using statistical features and forecasts future values with an LSTM model. Includes a Streamlit app for visualization.
This project focuses on implementing an unsupervised method for anomaly detection in time series data, particularly in vibratory signals. The main goal is to enhance early fault detection in industrial machines by using advanced statistical and machine learning algorithms. The developed system allows for continuous monitoring, improving maintenance planning, reducing operational costs, and preventing unexpected breakdowns.
requirements.txt
git clone https://github.com/Abdelhamid2c/anomaly-detection-in-time-series-based-on-statistical-features-and-forcasting.git
cd <your-repository-folder>
pip install -r requirements.txt
streamlit run Streamlit/PdM.py