☠ Software Defect Prediction

Machine learning model for 'Software Defect Prediction' using deep learning and based on Python & Tensorflow

Introduction

• Datasets

  Datasets for this project taken from PROMISE public dataset

• Models

  CNN, RNN, LSTM, Random Forest, and more

• Results

  Results directory

  Paper

  Results from running model(s) shown in reports folder

---

Project Structure

│datasets
├── processed
│   └── big_data1.csv
│   └── big_data2.csv
│   └── pc4.csv
│   └── pc3.csv
│   └── ...
├── raw
    └── ...

│references
└── README.md

│reports
├── figures
│   └── confussion matrix
│       └── random_forest.png
│       └── cnn.png
│       └── lstm.png
│   └── preprocess_balanced.png
│   └── preprocess_imbalanced.png
├── results
│   └── random_forest.txt
│   └── cnn.txt
│   └── lstm.txt
├── PAPER-Software Defect Prediction.pdf
└── preprocess.txt
└── PAPER-Software Defect Prediction
│src
├── models
│   ├── cnn.py
│   ├── lstm.py
│   ├── random_forest.py
├── main.py
└── preprocess.py

│README.md
│requirements.txt

---

Installation

💻 Windows 10 steps

Clone this repository or download it manual as a zip

$ git clone https://github.com/mhnaufal/Software-Defect-Finale.git

Open up cmd or Powershell (Powershell prefered) as Administrator and go to this repo directory

Create Python virtual environment:

$ python -m venv sddl-env

Run the virtual environment:

$ sddl-env/Scripts/activate

Install the library:

$ pip install -r requirements.txt

If above command result an error, run the cmd or Powershell as Administrator and then re run the above command

Run the models:

$ python src/main.py

or

$ python src/models/random_forest.py

Credits

Inspired by many other studies listed in here

License

MIT