🔤 TF-IDF + Traditional Machine Learning
Fast, interpretable baselines for text classification
📊 Feature Extraction
TF-IDF (Term Frequency - Inverse Document Frequency)
Converts text into numerical vectors by weighting words based on frequency and uniqueness. Common words (e.g., "the", "is") get low weights, while distinctive words get high weights.
Formula: TF-IDF(t,d) = TF(t,d) × IDF(t) = (term freq in doc) × log(total docs / docs with term)
🤖 Classification Algorithms
Naive Bayes
Probabilistic classifier based on Bayes theorem. Fast, efficient, works well with small datasets.
Logistic Regression
Linear model with sigmoid activation. Simple, interpretable, good baseline.
Support Vector Machine
Find optimal hyperplane for classification. Powerful for high-dimensional data.
Random Forest
Ensemble of decision trees. Robust, handles non-linearity, reduces overfitting.
Neural Network (MLP)
Multi-layer perceptron with hidden layers. Can learn complex patterns.
XGBoost
Gradient boosting algorithm. Often wins ML competitions. High performance.
💻 Full Implementation Code
Ready-to-run Python code with all 6 classifiers. Includes dataset download, TF-IDF extraction, training, and evaluation.
Features:
- ✅ Automatic dataset download from GitHub Pages
- ✅ TF-IDF feature extraction (5000 features, bigrams)
- ✅ 6 classifiers: Naive Bayes, Logistic, SVM, Random Forest, MLP, XGBoost
- ✅ Training & inference time measurements
- ✅ Classification reports & confusion matrices
- ✅ Visualizations saved as PNG files
Run in Google Colab
Open this notebook in Google Colab for interactive execution with free GPU access.
No Setup Required
All dependencies pre-installed
Interactive
Modify code and see results instantly
Save Results
Download plots and models
Note: The notebook will open in a new tab. You may need to sign in with your Google account.