透過對物理量的預測了解物體的品質狀態

We use conda to build our enviornment. The following will provide step-by-step instructions on how to set up the environment and run the code.

Installation environment / Execution environment / Package selection

Conda version: 23.1.0
Python version: 3.6.13
Main libraries used: scikit-learn, pandas, numpy, math, pickle, matplotlib

Environment setup

Unzip 113061_Source.zip and move into the folder.

unzip 113061_Source.zip
cd 113061_Source/

An environment can be created with all the Python dependencies.

conda env create -f environment.yml

Activate environment to run .py files.

conda activate 113061_Source

Execution method

1. Transfer CSV files to merged dataframe

Convert the CSV files under various resistor and voltage conditions into merged dataframe for easier reading.

python src/concatenation.py

2. Prepare training data

Divide the dataframe under each resistor and voltage condition into the first 50 timesteps and the following 3950 timesteps, while also separating the 10 experiments individually. Store them in a dictionary structure.

pyhon src/preprocessing.py

3. Train MLP model

python train.py

GengYunTien/mlpregressor