Stress Testing Anomaly Detection via Autoencoders with PyQT GUI

Setup:

• a pc-stress-monitor-dataset/ folder is required, situated in the root:

.
└── pc-stress-monitor-dataset/
    ├── cpu-loads-logs/
    │   ├── anomalous/
    │   ├── non-anomalous/
    │   └── anomaly-mapping.csv
    ├── cpu-temps-logs/
    │   ├── anomalous/
    │   ├── non-anomalous/
    │   └── anomaly-mapping.csv
    ├── pc-fans-logs/
    │   ├── anomalous/
    │   ├── non-anomalous/
    │   └── anomaly-mapping.csv
    ├── gpu-loads-logs/
    │   ├── anomalous/
    │   ├── non-anomalous/
    │   └── anomaly-mapping.csv
    ├── gpu-temps-logs/
    │   ├── anomalous/
    │   ├── non-anomalous/
    │   └── anomaly-mapping.csv
    └── gpu-fans-logs/
        ├── anomalous/
        ├── non-anomalous/
        └── anomaly-mapping.csv

each anomalous/ folders should have atleast one (1) appropriate csv logs without headers, but the first column must pertain to the time column (min time = 0, max time = 60) and the succeeding n columns are the features, where n > 0.

• models/ folder is optional as the application will create it on startup if it doesn't exist. An existing models/ folder can then be reused.
• OpenHardwareMonitorLib.dll is a dependency and must be present in the root. Can be downloaded here.
• Create a venv/ with python version >= 3.10.1, activate the virtual environment, then run pip install -r requirements.txt
• All is setup and ready to run main.py

Diagnosis demo