Machine Learning algorithms using pure Python; without libraries on various datasets (eg. UCI Machine Learning repository, Kaggle)
The main purpose of this repository is to understand and implement various Machine learning algorithms using pure Python; without using their existing library implementations (eg. from scikit-learn).
| Project | Dataset | Accuracy | Training samples/rows | Number of features/columns |
|---|---|---|---|---|
| Banknote authentication using Decision Trees | click here | 94.66% | 1372 | 4 |
| Diabetes prediction using Logistic Regression | click here | 78.78% | 768 | 8 |
| Diabetes prediction using Naive Bayes | click here | 77.06% | 768 | 8 |
| Iris plant species prediction using kNN | click here | 93.33% | 150 | 4 |