Saharkakavand/risk_prediction

automotive risk prediction code

risk_prediction

• quickstart
  • install
• package outline
  • scripts
  • data
  • tests

quickstart

install

• First install the Julia package AutoRisk.jl at https://github.com/wulfebw/AutoRisk.jl
• Next download or git clone this repository

package outline

The goal of this package, "risk_prediction", is to make it easy to predict automotive risks of datasets generated by AutoRisk.jl. It builds on AutoRisk.jl.

scripts

• This directory contains scripts for using AutoRisk to collect datasets (collection), for fitting models to those datasets (compression and prediction), and for visualizing datasets and scenarios (visualization).

data

• This directory holds data associated with the project such as generated datasets, log files, neural network files (useable from julia), snapshots and summaries (files saved by tensorflow during training), and visualizations (generally the output from the visualization scripts).

tests

• This directory contains tests for the python code in the scripts directory. To run the tests:

python run_tests.py