Added Random Forest Regressor as an additional prediction model.

DIRECTORY.md

@@ -471,6 +471,7 @@
   * [Geometry](geometry/geometry.py)
   * [Graham Scan](geometry/graham_scan.py)
   * [Jarvis March](geometry/jarvis_march.py)
+  * [Segment Intersection](geometry/segment_intersection.py)
   * Tests
     * [Test Graham Scan](geometry/tests/test_graham_scan.py)
     * [Test Jarvis March](geometry/tests/test_jarvis_march.py)

machine_learning/forecasting/run.py

@@ -13,8 +13,10 @@
 
 from warnings import simplefilter
 
+import matplotlib.pyplot as plt
 import numpy as np
 import pandas as pd
+from sklearn.ensemble import RandomForestRegressor
 from sklearn.preprocessing import Normalizer
 from sklearn.svm import SVR
 from statsmodels.tsa.statespace.sarimax import SARIMAX
@@ -78,6 +80,29 @@ def support_vector_regressor(x_train: list, x_test: list, train_user: list) -> f
     return float(y_pred[0])
 
 
+def random_forest_regressor(x_train: list, x_test: list, train_user: list) -> float:
+    """
+    Fourth method: Random Forest Regressor
+    Random Forest is an ensemble learning method for regression that operates
+    by constructing a multitude of decision trees at training time and outputting
+    the mean prediction of the individual trees.
+
+    It is more robust than a single decision tree and less prone to overfitting.
+    Good for capturing nonlinear relationships in data.
+
+    input : training data (date, total_event) in list of float
+    where x = list of set (date and total event)
+    output : list of total user prediction in float
+
+    >>> random_forest_regressor([[5,2],[1,5],[6,2]], [[3,2]], [2,1,4])
+    1.95
+    """
+    model = RandomForestRegressor(n_estimators=100, random_state=42)
+    model.fit(x_train, train_user)
+    prediction = model.predict(x_test)
+    return float(prediction[0])
+
+
 def interquartile_range_checker(train_user: list) -> float:
     """
     Optional method: interquatile range
@@ -120,6 +145,22 @@ def data_safety_checker(list_vote: list, actual_result: float) -> bool:
     return safe > not_safe
 
 
+def plot_forecast(actual, predictions):
+    plt.figure(figsize=(10, 5))
+    plt.plot(range(len(actual)), actual, label="Actual")
+    plt.plot(len(actual), predictions[0], "ro", label="Linear Reg")
+    plt.plot(len(actual), predictions[1], "go", label="SARIMAX")
+    plt.plot(len(actual), predictions[2], "bo", label="SVR")
+    plt.plot(len(actual), predictions[3], "yo", label="RF")
+    plt.legend()
+    plt.title("Data Safety Forecast")
+    plt.xlabel("Days")
+    plt.ylabel("Normalized User Count")
+    plt.grid(True)
+    plt.tight_layout()
+    plt.show()
+
+
 if __name__ == "__main__":
     """
     data column = total user in a day, how much online event held in one day,
@@ -155,8 +196,11 @@ def data_safety_checker(list_vote: list, actual_result: float) -> bool:
         ),
         sarimax_predictor(train_user, train_match, test_match),
         support_vector_regressor(x_train, x_test, train_user),
+        random_forest_regressor(x_train, x_test, train_user),
     ]
 
     # check the safety of today's data
     not_str = "" if data_safety_checker(res_vote, test_user[0]) else "not "
     print(f"Today's data is {not_str}safe.")
+
+    plot_forecast(train_user, res_vote)