A combination of lightweight, high specific strength, and good castability make magnesium alloys a promising engineering material for the automotive and aerospace industries. Vehicle weight reduction is one of the major means available to improve automotive fuel efficiency. High-strength steels, Aluminium (Al), and polymers are already being used to reduce weight significantly, but substantial additional reductions could be achieved by greater use of low-density magnesium (Mg) and its alloys. This project herein, therefore, relies on the use of machine learning, to assist in the development of A.I. to predict alloy compositions that are potentially useful for future metallic alloys. This study shows how a machine learning approach is able to offer acceptable precision predictions with respect to the main mechanical properties of metals.
u6734495_Comp4560.ipynb - contains the code in ipynb format and all the outputs from our code are shown in the file
u6734495_Comp4560.py - contains the code file in a .py format
The database for the project which was created from sctrach is present in Excel as well CSV format. (Contains data from MATWEB, CES material selection, research papers and journal articles)
matweb.py contains the code for scraping the data from MATWEB.
All the packages are imported in the code file. These are some important packages that needs to pre - installed
Tensorflow
XGBoost
tf_docs
For installing packages in python3 and ubuntu 18.04. Open command prompt and type the following one by one:
pip3 install tensorflow
pip3 install xgboost
pip3 install tensorflow_docs
For python2:
pip install tensorflow
pip install xgboost
pip install tensorflow_docs
- Samyak Jain
- Prateek Arora (u6742441)
- Bhumipat Chatlekhavanich (u6069393)
- Jia Ye (u5879731)
This project is licensed under the Australian National University License
- Professor Nick Birbillis - Deputy Dean College of Engineering & Computer Sciences
- Zhuoran (Randy) Zeng, Research Fellow Monash University, Victoria, Australia