CMPE 255 – Data Mining Spring 2023

Ethylene and Methane Detection using Machine Learning

This project uses machine learning to detect the presence of ethylene and methane in a gas mixture. The

project is divided into two parts:

1. Data preprocessing
2. Model training and evaluation

Data preprocessing

The first step is to preprocess the data.

This involves:

1. Reading the data from a CSV file
2. Dropping any rows with missing values
3. Splitting the data into training and test sets
4. Standardizing the features

Model training and evaluation

Once the data has been preprocessed, we can train and evaluate a model.

We will use a variety of machine learning models, including:

• Linear discriminant analysis (LDA)
• Quadratic discriminant analysis (QDA)
• Logistic regression

We have evaluated each model using accuracy, precision, recall, and F1 score.

Results

The best performing model is Logistic Regression, with an accuracy of 94% on the test set.

Conclusion

This project demonstrates the use of machine learning to detect the presence of ethylene and methane in a gas mixture. The results show that LDA is a promising model for this task.