Machine Learning Model Predicting
Strength and Conductivity of
Copper alloys (POONGSAN data) ver.0.5b

Our website provides a machine learning model to predict Vickers hardness and electrical conductivity of copper alloys. The purpose of this website is to demonstrate the excellence of our machine learning model. Therefore, detailed information on the material synthesis process is omitted.
You can also find a machine learning model for predicting thermoelectric properties of BiTe-based materials at our website.

Please follow the steps below to predict the material properties.

Step 1: Choose Material Composition and Producing Method

SHT stands for Solution Heat Treatment. In some methods, aging is performed twice (1st aging/aging). If 1st aging temp and time are zero, it means that aging is performed only once (no 1st aging).

Scatter Plot: Experiments by Poongsan (Expt)
Plot 1: Machine Prediction (Pred 1)
Plot 2: Machine Prediction (Pred 2)
Show Options

Step 2: Check the Material Properties (Our Model)

In the plot, the filled region means 95% CI (confidence interval) of the mean. However, because the synthesis is rarely repeated in this example, the CI has no significant meaning.

Material Properties

Compare: Prediction by a Standard Neural Network

The following shows a prediction by a fully connected neural network that has a similar size to our model (about 72,000 weights).

Material Properties

Methods

Contacts

Powered by