https://github.com/duzzisantos/software-estimation-training
https://github.com/duzzisantos/software-estimation-training
Last synced: about 2 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/duzzisantos/software-estimation-training
- Owner: duzzisantos
- Created: 2024-09-29T08:41:59.000Z (7 months ago)
- Default Branch: main
- Last Pushed: 2024-11-16T19:37:08.000Z (5 months ago)
- Last Synced: 2025-01-16T00:57:29.088Z (3 months ago)
- Language: Python
- Size: 47.9 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# About
This is a backend application that helps to address one of the classical problems in Information Systems: Software Project Estimation Bias. It seeks to collect data from a client application which logs in individual work logs for sub tasks that make up a bigger task.
As task/work time logs are recorded, they are passed through Machine Learning tools like TensorFlow, Numpy and Scikit-Learn which are notably applied here to perform time-series forecasts using historical work/task time logs as well as random sampling and probability distribution functions.
The trained time series data are stored as batch data, to enable client view trained data from the past - with the view of detecting anomalies or deviations in the prediction as well as investigating the causes (which might be outside of the application data's scope. Think - what if there was a layoff, or staff were stationed to other projects - thus prolonging the time required to deliver well-known tasks?). To build this data training model, the Long-Term Short-Term Memory recurrent neural network (RNN) is applied, that way we can predict long-term non-linear series.
The PERT analysis is passed through a triangular randomization function and further passed through a Monte Carlo simulation to yield numerous amounts of simulations (10000), a mean duration, standard deviation and duration commonest in the 90th percentile.
## Tools used
- Python
- FastAPI
- MongoDB
- TensorFlow
- Scikit-Learn
- Pandas
- Numpy
- Mathplotlib
## Objectives
- To generate experimental data for a research paper addressing the issue of software project time estimation bias.
- To compare previous academic research outcomes on this topic -
using results generated from this system.
- To have a system that guides developers and project managers toward making more-informed, objective decisions with regards to estimating effort and time expended in building software.
## Process flow
See flow chart here:
https://ibb.co/XpNCg7Q