• 

  Confusion matrix' heatmap

Performance

• F1 Score - Micro: 0.780331373597007
• F1 Score - Macro: 0.5981815091402464
• F1 Score - Weighted: 0.7443181899051569
• Cohen’s kappa score: 0.6355790715803891

Inspiration

Alexis Nolin-Lapalme's intro to Machine Learning

One of the biggest problems in machine learning is overfitting, but most of the time this won’t happen thanks to the random forest classifier. If there are enough trees in the forest, the classifier won’t overfit the model.

Source: https://builtin.com/data-science/random-forest-algorithm

[...] the Random Forest (RF) algorithm showed superior accuracy comparatively. Of the 17 studies where it was applied, RF showed the highest accuracy in 9 of them, i.e., 53%.

Source: https://bmcmedinformdecismak.biomedcentral.com/articles/10.1186/s12911-019-1004-8

What it does

Classify different diseases using a multi-label classification model based on the microorganisms in the gut microbiome.

How we built it

• Python
• scikit-learn - Machine Learning
• seaborn - Heatmaps
• pandas - Data Analysis

Challenges we ran into

• Separating 'Disease-2' from 'Healthy' samples

Accomplishments that we're proud of

• Built a functional classifier with minimal code, minimal complexity, minimal performance cost

What we learned

• Random Forest Classifier
• Using scikit-learn for Machine Learning

What's next?

• Use sklearn-onnx to export our model in a standard format
• Include potential future q-values -> FDR
• Tuning to separate disease-2 from our healthy samples

Built with

• Python
• scikit-learn - Machine Learning
• seaborn - Heatmaps
• pandas - Data Analysis

Built With

• pandas
• python
• scikit-learn
• seaborn