Thyroid Disease prediction

In this project we use Machine Learning (ML) models to predict if a person has thyroid or not based on the data given.
This project also has data visualization techniques.

Dataset

The data used is available in kaggle: Thyroid Dataset

Packages used:

• scikit-learn
• pandas
• seaborn
• matplot-lib
• plotly
• imbalanced-learn

Execute this code in your terminal before you start jupyter notebook:

pip install -r requirements.txt

This above code install all the required packages to run the model NOTE: This above code is required only if you are running in your computer not in colab

Information about model:

In this model, I will predict two thyroid conditions and if a person have thyroid or not, and also predict the hyperthyroid condition. So this model is muliclass classification model. Algorithms used:

1. Decision Trees
2. Logistic Regression
3. Random Forest
4. Support Vector Machine
5. Feed Forward (Multi layer perceptron)

Results Achieved:

Model	Accuracy
Decision Tree	99.37%
Logistic Regression	91.47%
Random Forest	99.90%
SVM	94.94%
Feed Forward	98.50%