Modelo LinearRegression v0.1.0

README.md

@@ -0,0 +1,84 @@
+---
+license: apache-2.0
+library_name: sklearn
+pipeline_tag: tabular-regression
+tags:
+  - linear-regression
+  - tabular
+  - regression
+  - pastelerias
+  - ventas
+---
+
+# pastelerias-sales-predictor
+
+## Información del Modelo
+- **Tipo**: Regresión Lineal Múltiple (OLS)
+- **Framework**: scikit-learn
+- **Autor**: gusdelact
+- **Fecha de entrenamiento**: 2026-06-24T22:00:39.750756
+- **Formato**: joblib
+
+## Uso Previsto
+- **Tarea**: Regresión — predecir ventas mensuales de pastelerías
+- **Variable target**: `Monthly sales`
+- **Contexto**: Planificación de expansión de negocio de pastelerías
+
+## Datos de Entrenamiento
+- **Fuente**: [gusdelact/pasteleria](https://huggingface.co/datasets/gusdelact/pasteleria)
+- **Dataset curado**: [gusdelact/pastelerias-curated](https://huggingface.co/datasets/gusdelact/pastelerias-curated)
+- **Samples**: 10
+- **Features**: 2
+
+## Ecuación del Modelo
+
+```
+Monthly sales = 65.32
+              + 41.51 × Floor space
+              - 0.3409 × Distance to station
+```
+
+## Métricas de Evaluación (Leave-One-Out CV, N=10)
+
+| Métrica | Valor |
+|---------|-------|
+| RMSE | 27.02 |
+| MAE | 22.26 |
+| R² | 0.9042 |
+
+**Umbral de aceptación**: RMSE ≤ 50 → ✅ Cumplido
+
+## Cómo Usar
+
+```python
+import joblib
+import pandas as pd
+from huggingface_hub import hf_hub_download
+
+# Descargar modelo
+model_path = hf_hub_download("gusdelact/pastelerias-linear-regression", "model.joblib")
+model = joblib.load(model_path)
+
+# Predecir
+X_new = pd.DataFrame({
+    "Floor space of the shop": [8],
+    "Distance to the nearest station": [150]
+})
+prediction = model.predict(X_new)
+print(f"Ventas predichas: {prediction[0]:.0f} unidades/mes")
+```
+
+## Interpretación de Coeficientes
+- **+1 unidad de Floor space** → +41.5 ventas/mes
+- **+1 metro de distancia a estación** → -0.34 ventas/mes
+- **Base (intercepto)**: 65.3 ventas/mes
+
+## Limitaciones
+- Entrenado con solo 10 observaciones — intervalos de confianza amplios
+- Solo válido dentro del rango observado (Floor: 5-10, Distance: 0-330)
+- No captura relaciones no lineales ni interacciones
+- Datos de una región específica (Japón)
+
+## Fundamento Teórico
+Ver `notes/03_design_modeling.md` en el repositorio del proyecto para la fundamentación
+completa con citas a ESL §3.2 (OLS), ISLP §5.1.4 (LOO-CV) y FES §3.4.1 (CV con N pequeño).

metrics.json

@@ -0,0 +1,11 @@
+{
+  "rmse_loo_cv": 27.016858562584517,
+  "mae_loo_cv": 22.25680200197573,
+  "r2_loo_cv": 0.9042106984038373,
+  "rmse_train": 20.42793704708016,
+  "r2_train": 0.945235852681711,
+  "rmse_threshold": 50.0,
+  "threshold_met": true,
+  "n_samples": 10,
+  "cv_method": "Leave-One-Out"
+}

model.joblib

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8fea73435e71f3af3f103d732dc4e3166845d802ee0833449fe074be3da5dd5a
+size 951

model_info.json

@@ -0,0 +1,30 @@
+{
+  "model_type": "LinearRegression",
+  "framework": "scikit-learn",
+  "target": "Monthly sales",
+  "feature_order": [
+    "Floor space of the shop",
+    "Distance to the nearest station"
+  ],
+  "intercept": 65.32391638894813,
+  "coefficients": {
+    "Floor space of the shop": 41.5134782564385,
+    "Distance to the nearest station": -0.3408826856636193
+  },
+  "metrics": {
+    "rmse_loo_cv": 27.016858562584517,
+    "r2_loo_cv": 0.9042106984038373
+  },
+  "input_ranges": {
+    "Floor space of the shop": {
+      "min": 5,
+      "max": 10,
+      "type": "int"
+    },
+    "Distance to the nearest station": {
+      "min": 0,
+      "max": 330,
+      "type": "int"
+    }
+  }
+}

train_metadata.json

@@ -0,0 +1,23 @@
+{
+  "model_type": "LinearRegression",
+  "framework": "scikit-learn",
+  "timestamp": "2026-06-24T22:00:39.750756",
+  "n_samples": 10,
+  "n_features": 2,
+  "features": [
+    "Floor space of the shop",
+    "Distance to the nearest station"
+  ],
+  "target": "Monthly sales",
+  "params": {
+    "fit_intercept": true
+  },
+  "intercept": 65.32391638894813,
+  "coefficients": {
+    "Floor space of the shop": 41.5134782564385,
+    "Distance to the nearest station": -0.3408826856636193
+  },
+  "condition_index": 36.83804573206135,
+  "ols_r_squared": 0.945235852681711,
+  "ols_adj_r_squared": 0.929588953447914
+}