feat(website): update for v0.3.0 with new algorithms

- Update version references from 0.2.0 to 0.3.0
- Add 6 new trainers to trainers page (LightGBM, CatBoost, K-Means, DBSCAN, Isolation Forest, One-Class SVM)
- Organize trainers by category with section headers
- Update stats: 7+ ML Models -> 15+ ML Models
- Update features description to include new algorithms
- Update sidebar with trainer subsections
- Update fallback release data for v0.3.0
- Add v0.3.0 to version selector

Author: codeMaestro78

CHANGELOG.md

@@ -8,18 +8,22 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 ## [0.3.0] - 2025-12-14
 
 ### Added
+
 - **Gradient Boosting Algorithms**
+
   - LightGBM trainer with classification/regression support
   - CatBoost trainer with automatic categorical feature handling
   - Early stopping, feature importance, and native model exports
 
 - **Clustering Algorithms**
+
   - K-Means trainer with silhouette, calinski-harabasz, davies-bouldin metrics
   - DBSCAN trainer with automatic noise detection
   - Optimal cluster number finder (elbow method) for K-Means
   - Optimal eps finder for DBSCAN
 
 - **Anomaly Detection Algorithms**
+
   - Isolation Forest trainer for efficient outlier detection
   - One-Class SVM trainer with RBF kernel support
   - Anomaly scoring and threshold tuning capabilities
@@ -31,6 +35,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
   - Added anomaly/isolation_forest.json and anomaly/one_class_svm.json
 
 ### Changed
+
 - Updated trainer registry with 6 new algorithm types
 - Extended model type support: classification, regression, clustering, anomaly_detection
 
@@ -39,7 +44,9 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 ## [0.2.0] - 2025-12-07
 
 ### Added
+
 - **Documentation**
+
   - Created comprehensive `docs/` folder with getting-started guide, installation instructions
   - Added `examples/` folder with sample configs for all model types (RF, XGB, Logistic, DNN)
   - Added example tuning configurations for hyperparameter optimization
@@ -54,59 +61,70 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
   - Added shared fixtures and configuration in `conftest.py`
 
 ### Changed
+
 - Cleaned up project structure for production deployment
 - Updated `.gitignore` to exclude generated artifacts properly
 - Improved GridSearchTuner to handle `n_trials` parameter gracefully
 
 ### Removed
+
 - Hotel-specific datasets and configurations (moved to generic examples)
 - Test experiment artifacts from repository
 - Empty `scripts/` directory
 
 ### Fixed
+
 - GridSearchTuner now properly filters out unsupported `n_trials` parameter
 - Configuration files now use generic `data/your_data.csv` paths
 
 ## [0.1.0] - 2025-12-03
 
 ### Added
+
 - **CLI Training Pipeline**
+
   - Train ML models (Logistic Regression, SVM, Random Forest, XGBoost)
   - Train DL models (TensorFlow DNN, CNN, RNN/LSTM/GRU)
   - Configuration-driven training via JSON/YAML files
   - Parameter overrides from command line
 
 - **Hyperparameter Tuning**
+
   - Grid Search for exhaustive parameter search
   - Random Search for large parameter spaces
   - Bayesian Optimization via Optuna for intelligent search
   - Cross-validation support
   - Auto-train best model after tuning
 
 - **Model Explainability**
+
   - SHAP (SHapley Additive exPlanations) for global/local explanations
   - LIME (Local Interpretable Model-agnostic Explanations)
   - Feature importance visualization
   - Instance-level explanations
 
 - **Data Preprocessing Pipeline**
+
   - Scaling: StandardScaler, MinMaxScaler, RobustScaler
   - Normalization: L1, L2, Max norm
   - Encoding: LabelEncoder, OneHotEncoder, OrdinalEncoder
   - Feature Selection: SelectKBest, RFE, VarianceThreshold
   - Pipeline support for chaining preprocessors
 
 - **Experiment Tracking**
+
   - Automatic experiment logging
   - Run comparison and filtering
   - Export to CSV
   - Mini-MLflow style tracking
 
 - **Model Export**
+
   - ML models: Pickle, Joblib, ONNX
   - DL models: SavedModel, H5
 
 - **Interactive Terminal UI (TUI)**
+
   - Train models with guided interface
   - Evaluate saved models
   - Browse experiment history
@@ -118,6 +136,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
   - Easy extension for custom models
 
 ### Technical Details
+
 - Python 3.8+ support
 - Type hints throughout codebase
 - Rich CLI output with colors and tables

mlcli/trainers/anomaly/isolation_forest_trainer.py

@@ -143,8 +143,7 @@ def train(
         self.is_trained = True
 
         logger.info(
-            f"Training complete. Detected {n_anomalies} anomalies "
-            f"({anomaly_ratio:.2%} of data)"
+            f"Training complete. Detected {n_anomalies} anomalies " f"({anomaly_ratio:.2%} of data)"
         )
 
         return self.training_history
@@ -189,9 +188,7 @@ def _compute_detection_metrics(
 
         return metrics
 
-    def evaluate(
-        self, X_test: np.ndarray, y_test: np.ndarray = None
-    ) -> Dict[str, float]:
+    def evaluate(self, X_test: np.ndarray, y_test: np.ndarray = None) -> Dict[str, float]:
         """
         Evaluate anomaly detection on test data.
 

mlcli/trainers/anomaly/one_class_svm_trainer.py

@@ -190,9 +190,7 @@ def _compute_detection_metrics(
 
         return metrics
 
-    def evaluate(
-        self, X_test: np.ndarray, y_test: np.ndarray = None
-    ) -> Dict[str, float]:
+    def evaluate(self, X_test: np.ndarray, y_test: np.ndarray = None) -> Dict[str, float]:
         """
         Evaluate anomaly detection on test data.
 

mlcli/trainers/catboost_trainer.py

@@ -227,9 +227,7 @@ def get_feature_importance(self) -> Dict[str, np.ndarray]:
             "prediction_values_change": self.model.get_feature_importance(
                 type="PredictionValuesChange"
             ),
-            "loss_function_change": self.model.get_feature_importance(
-                type="LossFunctionChange"
-            ),
+            "loss_function_change": self.model.get_feature_importance(type="LossFunctionChange"),
         }
 
     def save(self, save_dir: Path, formats: List[str]) -> Dict[str, Path]:

mlcli/trainers/clustering/dbscan_trainer.py

@@ -120,9 +120,7 @@ def train(
 
         self.is_trained = True
 
-        logger.info(
-            f"Training complete. Found {n_clusters} clusters and {n_noise} noise points"
-        )
+        logger.info(f"Training complete. Found {n_clusters} clusters and {n_noise} noise points")
         if train_metrics.get("silhouette") is not None:
             logger.info(f"Silhouette score: {train_metrics['silhouette']:.4f}")
 
@@ -139,14 +137,9 @@ def _get_cluster_sizes(self, labels: np.ndarray) -> Dict[str, int]:
             Dictionary mapping cluster id to size
         """
         unique, counts = np.unique(labels, return_counts=True)
-        return {
-            f"cluster_{int(k)}" if k != -1 else "noise": int(v)
-            for k, v in zip(unique, counts)
-        }
+        return {f"cluster_{int(k)}" if k != -1 else "noise": int(v) for k, v in zip(unique, counts)}
 
-    def _compute_clustering_metrics(
-        self, X: np.ndarray, labels: np.ndarray
-    ) -> Dict[str, float]:
+    def _compute_clustering_metrics(self, X: np.ndarray, labels: np.ndarray) -> Dict[str, float]:
         """
         Compute clustering evaluation metrics.
 
@@ -171,9 +164,7 @@ def _compute_clustering_metrics(
             metrics["calinski_harabasz"] = float(
                 calinski_harabasz_score(X_filtered, labels_filtered)
             )
-            metrics["davies_bouldin"] = float(
-                davies_bouldin_score(X_filtered, labels_filtered)
-            )
+            metrics["davies_bouldin"] = float(davies_bouldin_score(X_filtered, labels_filtered))
         else:
             metrics["silhouette"] = None
             metrics["calinski_harabasz"] = None
@@ -272,9 +263,7 @@ def fit_predict(self, X: np.ndarray) -> np.ndarray:
         self.train(X)
         return self._labels
 
-    def find_optimal_eps(
-        self, X: np.ndarray, k: int = None, plot: bool = False
-    ) -> float:
+    def find_optimal_eps(self, X: np.ndarray, k: int = None, plot: bool = False) -> float:
         """
         Find optimal eps using k-distance graph method.
 

mlcli/trainers/clustering/kmeans_trainer.py

@@ -57,9 +57,7 @@ def __init__(self, config: Optional[Dict[str, Any]] = None):
         default_params = self.get_default_params()
         self.model_params = {**default_params, **params}
 
-        logger.info(
-            f"Initialized KMeansTrainer with n_clusters={self.model_params['n_clusters']}"
-        )
+        logger.info(f"Initialized KMeansTrainer with n_clusters={self.model_params['n_clusters']}")
 
     def train(
         self,
@@ -121,9 +119,7 @@ def train(
 
         return self.training_history
 
-    def _compute_clustering_metrics(
-        self, X: np.ndarray, labels: np.ndarray
-    ) -> Dict[str, float]:
+    def _compute_clustering_metrics(self, X: np.ndarray, labels: np.ndarray) -> Dict[str, float]:
         """
         Compute clustering evaluation metrics.
 
@@ -227,9 +223,7 @@ def get_cluster_centers(self) -> np.ndarray:
 
         return self.model.cluster_centers_
 
-    def find_optimal_k(
-        self, X: np.ndarray, k_range: range = range(2, 11)
-    ) -> Dict[str, Any]:
+    def find_optimal_k(self, X: np.ndarray, k_range: range = range(2, 11)) -> Dict[str, Any]:
         """
         Find optimal number of clusters using elbow method and silhouette analysis.
 

website/src/app/about/page.tsx

@@ -19,7 +19,7 @@ const values = [
     icon: Target,
     title: 'Accurate Results',
     description:
-      'Built on proven ML libraries like scikit-learn, XGBoost, LightGBM, and TensorFlow.',
+      'Built on proven ML libraries like scikit-learn, XGBoost, LightGBM, CatBoost, and TensorFlow.',
   },
   {
     icon: Users,

website/src/app/docs/quickstart/page.tsx

@@ -27,7 +27,7 @@ export default function QuickstartPage() {
       <p>Verify the installation:</p>
       <CodeBlock
         code={`mlcli --version
-# mlcli v0.2.0`}
+# mlcli v0.3.0`}
         language="bash"
       />