INDUSTRY:
HEALTH
Course:
DATA MINING
YEAR:
2024
Project Heart
about.
The main goal of this project was to create models that predict whether people are likely to have heart disease based on different health factors. We had a dataset with various features like age, blood pressure, and cholesterol levels. The challenge was using these features to predict whether someone has heart disease accurately.
challenge.
The main challenge of this project was to work with a dataset that combined both numerical and categorical health indicators while ensuring the integrity and usability of the data for predictive modelling. Handling outliers, zero or negative values, and inconsistent data types required careful preprocessing to avoid skewing the results. Another critical focus was selecting the most relevant features and designing models that could accurately classify heart failure risk while remaining interpretable.
result.
The analysis did a great job of predicting heart disease using various health indicators. The feature selection process helped by focusing on the most relevant features, which improved the models. Random Forest was the best for predicting heart disease because it handled complex data relationships well. Clustering gave us valuable insights but showed that we need a more detailed approach to separate different risk levels.
The results highlight the importance of detecting risk factors early, like high cholesterol and unusual blood pressure levels. Some limitations of this study are the dataset size and possible biases in the feature selection process. For future work, using larger datasets or trying more advanced algorithms like neural networks can help enhance predictive accuracy.
