https://github.com/slxiao/prober
:mag_right: Software bug predictor based on machine learning algorithm
https://github.com/slxiao/prober
ai bug-prediction machine-learning software-testing tensorflow
Last synced: 2 months ago
JSON representation
:mag_right: Software bug predictor based on machine learning algorithm
- Host: GitHub
- URL: https://github.com/slxiao/prober
- Owner: slxiao
- Created: 2019-11-23T07:01:36.000Z (almost 6 years ago)
- Default Branch: master
- Last Pushed: 2019-12-02T01:24:52.000Z (almost 6 years ago)
- Last Synced: 2025-07-11T23:03:08.483Z (3 months ago)
- Topics: ai, bug-prediction, machine-learning, software-testing, tensorflow
- Language: Java
- Homepage:
- Size: 31.3 KB
- Stars: 1
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Introduction
`prober` is an automated software bug predictor using machine learning techniques. Specifically, `proper` trains prediction model based on open source datasets, and then predicts whether some code has a bug or not. The prediction is based on various code features, such as McCabe features, Halstead features, etc.
# How it works
- traning data is from [open source dataset](./prober/data.csv)
- training algorithm is based on `tensorflow` and `keras`
- features extraction algorithm is specifc for each feature and each programming language. For java, the feature extraction algorithm can be found in [java](./java)
# How to use
## traning
Install dependencies:
```shell
pip install -r requirements.txt
```
Start traning:
```shell
python -m prober.prober
```
Extract java features:
```shell
gradle build && gradle execute
```
# LISENSE
MIT