{
"cells": [
{
"cell_type": "markdown",
"id": "8ab79e2c-d31e-48a4-bc76-97e80836cac9",
"metadata": {},
"source": [
"# Entscheidungsbäume\n",
"\n",
"Ein Entscheidungsbaum ist ein Klassifikationsverfahren, welches anhand von zu prüfenden Kriterien eine Klassifikation vornimmt. Dabei werden nacheinander unterschiedliche Kriterien durch ja-nein-Entscheidungen geprüft, bis eine Klassenzugehörigkeit festgestellt werden kann. \n",
"\n",
"Der Entscheidungsbaum wird gemäß der verwendeten Trainingsdaten aufgebaut. Die Bewertung seiner Genauigkeit erfolgt unter Benutzung von Testdaten. Ein geeigneter Entscheidungsbaum hat eine hohe Klassifikationsgenauigkeit, bei möglichst wenig auszuführenden Kriterientests.\n",
"\n",
"Die folgende Abbildung zeigt die grafische Darstellung eines einfachen Entscheidungsbaumes: \n",
"
\n",
"\n",
"--------------\n",
"\n",
"Matzka beschreibt Entscheidungsbäume in seinem Buch auf den Seiten 104-107.\n",
"\n",
"Folgende Webseiten zeigen die Implementierung mit Python, pandas und sklearn: \n",
"https://martin-grellmann.de/entscheidungsbaum-klassifikation-und-ihre-python-implementierung \n",
"https://www.tecislava.com/blog/decision-trees \n",
"https://www.w3schools.com/python/python_ml_decision_tree.asp"
]
},
{
"cell_type": "markdown",
"id": "f94d05e2-4983-4678-8cfa-6c9f8d0fe238",
"metadata": {},
"source": [
"Das folgende Beispiel zeigt die Nutzung von Entscheidungsbäumen am Beispiel des Schwertlilien-Datensatzes."
]
},
{
"cell_type": "markdown",
"id": "2c40aef4-e179-4a66-bc94-2e867d8e97f5",
"metadata": {},
"source": [
"## Eigene Übungsaufgabe\n",
"\n",
"Wir verwenden den Wein-Datensatz('data/Wine.csv'): https://www.geeksforgeeks.org/wine-dataset/ \n",
"Dieser enthält drei Klassen Wein mit insgesamt 13 Merkmalen und 178 Datensätzen. Für einen Entscheidungsbaum-Klassifikator ist folgendes zu untersuchen: \n",
"**Abhängigkeit der Genauigkeit von der Auswahl der Merkmale in den Testdaten bei vorgegebener maximaler Baumtiefe = 6**\n",
"\n",
"\n",
"### Was lässt sich über den Datensatz mithilfe der Entscheidungsbäume herausfinden?\n",
"\n",
"**Problem:**\n",
"Es wäre eigentlich Fachwissen nötig, um gute Klassifikationsmerkmale auszuwählen.\n",
"\n",
"**Lösungsansatz:**\n",
"Durch zufälliges Probieren von Merkmalskombinationen diejenigen identifizieren, die vielversprechende Ergebnisse liefern.\n"
]
},
{
"cell_type": "code",
"execution_count": 3,
"id": "89f16cdf-810b-47fc-812f-7ac8be604de3",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Umfang des Datensatzes: (178, 15)\n",
"---------------\n",
"neues Ergebnis:\n",
"Accuracy: 0.9662921348314607\n",
"Spalten:\n",
"[ 7 13 2 10]\n",
"Random State Daten:\n",
"439804\n",
"Random State Baum:\n",
"3815878\n",
"---------------\n",
"---------------\n",
"neues Ergebnis:\n",
"Accuracy: 0.9775280898876404\n",
"Spalten:\n",
"[13 7 10 3]\n",
"Random State Daten:\n",
"181513\n",
"Random State Baum:\n",
"3734753\n",
"---------------\n",
"---------------\n",
"neues Ergebnis:\n",
"Accuracy: 0.9887640449438202\n",
"Spalten:\n",
"[12 13 7 10]\n",
"Random State Daten:\n",
"871491\n",
"Random State Baum:\n",
"1337074\n",
"---------------\n",
"---------------\n",
"neues Ergebnis:\n",
"Accuracy: 1.0\n",
"Spalten:\n",
"[13 7 6 10]\n",
"Random State Daten:\n",
"2080615\n",
"Random State Baum:\n",
"858239\n",
"---------------\n"
]
},
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"# Numpy für Zufallszahlen\n",
"import numpy as np\n",
"\n",
"\n",
"\n",
"from sklearn.model_selection import train_test_split\n",
"from sklearn.tree import DecisionTreeClassifier, plot_tree\n",
"import matplotlib.pyplot as plt\n",
"from sklearn import metrics\n",
"import pandas as pd\n",
"\n",
"# Speichervariable für die maximal erreichte Genauigkeit\n",
"max_accuracy=0.95\n",
"\n",
"df = pd.read_csv('data/Wine.csv', sep=',')\n",
"# print(df.head())\n",
"print(\"Umfang des Datensatzes:\", df.shape)\n",
"\n",
"# Training und Genauigkeitsbestimmung in einer Endlosschleife mit zufälligen Spalten\n",
"while True:\n",
" # Auswahl der zufälligen Merkmalsspalten \n",
" randnum = np.random.randint(1,14,4)\n",
" \n",
" # Zufällige Merkmalsspalten (Features - X) und Zielwertspalte (Target- y) bereitstellen\n",
" X = df.iloc[:,randnum] \n",
" y = df['target']\n",
"\n",
" # Trainingsdaten und Testdaten erzeugen >>>>> test_size=0.5\n",
" # Wegen des großen Einflusses des random state wird auch hier eine Zufallszahl zwischen 1 und 4e6 verwendet\n",
" randstate_data = np.random.randint(1,4000000)\n",
" X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.5, random_state=randstate_data) \n",
"\n",
" # Entscheidungsbaum erzeugen - auch hier Verwendung einer Zufallszahl\n",
" randstate_tree = np.random.randint(1,4000000)\n",
" clf = DecisionTreeClassifier(max_depth=6, random_state=randstate_tree) \n",
" clf.fit(X_train, y_train) # Entscheidungsbaum anhand der Trainingsdaten aufbauen\n",
" y_pred = clf.predict(X_test) # mit dem Entscheidungsbaum klassifizieren\n",
"\n",
" # erreichte Genauigkeit bestimmen\n",
" accuracy = metrics.accuracy_score(y_test, y_pred)\n",
"\n",
" # nicht jedes Ergebnis soll angezeigt werden, falls eine Genauigkeit größer \n",
" #der zuvor maximal erreichten Genauigkeit bestimmt wird, sollen die Genauigkeit, \n",
" # die Merkmalsspalten, und die random States ausgegeben werden.\n",
" if accuracy > max_accuracy:\n",
" print('---------------')\n",
" print('neues Ergebnis:')\n",
" print(f\"Accuracy: {accuracy}\")\n",
" print('Spalten:')\n",
" print(randnum)\n",
" print('Random State Daten:')\n",
" print(randstate_data)\n",
" print('Random State Baum:')\n",
" print(randstate_tree)\n",
" print('---------------')\n",
" \n",
" max_accuracy=accuracy\n",
" # Abbruch der Schleife bei 99,9 % Genauigkeit \n",
" if accuracy > 0.999:\n",
" break\n",
"\n",
"# --- optional den Baum darstellen ---\n",
"plt.figure(figsize=(10, 5))\n",
"y = y.astype({'target':'string'}) # y in String konvertieren, damit der Plot funktioniert\n",
"plot_tree(clf, feature_names=X.columns, class_names=sorted(y.unique()), filled=True, rounded=True)\n",
"plt.show()\n",
"\n"
]
},
{
"cell_type": "markdown",
"id": "9c7db037-de73-4a01-a5fe-4448a0ebe4c7",
"metadata": {},
"source": [
"## Erkenntnisse\n",
"\n",
"- Spalte 13 taucht in allen Ergebnissen auf\n",
"- Spalte 10 taucht in allen Ergebnissen auf\n",
"- Spalte 7 taucht in allen Ergebnissen auf\n",
"- Spalte 2,3,6, und 12 tauchen jeweils einmal auf\n",
"\n",
"Relevant sind demnach wahrscheinlich:\n",
"- color_intensity\n",
"- flavanoids\n",
"- proline\n",
"\n",
"**Klassifikationsversuch mit den mehrfach auftauchenden 3 Kriterien --> Schwankungsbreite der Genauigkeit darstellen**\n",
"\n",
"**Klassifikationsversuch mit 3 Kriterien, die jeweils nur einmal auftauchen --> Schwankungsbreite der Genauigkeit darstellen**\n",
"\n",
"**Fragestellung:** \n",
"Ist die Klassifikation mit den 3 besten Kriterien wirklich besser oder wurde nur eine günstige Random-State Variable gewählt?\n"
]
},
{
"cell_type": "code",
"execution_count": 18,
"id": "3619cf74-a5e3-4427-9e5e-fed44476e1b3",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Umfang des Datensatzes: (178, 15)\n"
]
}
],
"source": [
"# Numpy für Zufallszahlen\n",
"import numpy as np\n",
"\n",
"from sklearn.model_selection import train_test_split\n",
"from sklearn.tree import DecisionTreeClassifier, plot_tree\n",
"import matplotlib.pyplot as plt\n",
"from sklearn import metrics\n",
"import pandas as pd\n",
"\n",
"df = pd.read_csv('data/Wine.csv', sep=',')\n",
"# print(df.head())\n",
"print(\"Umfang des Datensatzes:\", df.shape)\n",
"\n",
"# 2000 Durchläufe\n",
"accuracy_vector_gut = np.zeros((2000,1))\n",
"for i in range(1,2000):\n",
" # Auswahl der oben bestimmten Merkmalsspalten \n",
" randnum = [13,10,7]\n",
" \n",
" X = df.iloc[:,randnum] \n",
" y = df['target']\n",
"\n",
" # Trainingsdaten und Testdaten erzeugen >>>>> test_size=0.5\n",
" # Wegen des großen Einflusses des random state wird auch hier eine Zufallszahl zwischen 1 und 4e6 verwendet\n",
" randstate_data = np.random.randint(1,4000000)\n",
" X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.5, random_state=randstate_data) \n",
"\n",
" # Entscheidungsbaum erzeugen - auch hier Verwendung einer Zufallszahl\n",
" randstate_tree = np.random.randint(1,4000000)\n",
" clf = DecisionTreeClassifier(max_depth=6, random_state=randstate_tree) \n",
" clf.fit(X_train, y_train) # Entscheidungsbaum anhand der Trainingsdaten aufbauen\n",
" y_pred = clf.predict(X_test) # mit dem Entscheidungsbaum klassifizieren\n",
"\n",
" # erreichte Genauigkeit bestimmen\n",
" accuracy_vector_gut[i] =[metrics.accuracy_score(y_test, y_pred)]\n",
"\n",
" \n",
" "
]
},
{
"cell_type": "code",
"execution_count": 19,
"id": "b72bb2e7-e6b4-49dd-be96-3d678914115f",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Umfang des Datensatzes: (178, 15)\n"
]
}
],
"source": [
"# Numpy für Zufallszahlen\n",
"import numpy as np\n",
"\n",
"from sklearn.model_selection import train_test_split\n",
"from sklearn.tree import DecisionTreeClassifier, plot_tree\n",
"import matplotlib.pyplot as plt\n",
"from sklearn import metrics\n",
"import pandas as pd\n",
"\n",
"df = pd.read_csv('data/Wine.csv', sep=',')\n",
"# print(df.head())\n",
"print(\"Umfang des Datensatzes:\", df.shape)\n",
"\n",
"# 2000 Durchläufe\n",
"accuracy_vector_schlecht = np.zeros((2000,1))\n",
"for i in range(1,2000):\n",
" # Auswahl der beliebiger anderer Merkmalsspalten \n",
" randnum = [2,3,12]\n",
" \n",
" X = df.iloc[:,randnum] \n",
" y = df['target']\n",
"\n",
" # Trainingsdaten und Testdaten erzeugen >>>>> test_size=0.5\n",
" # Wegen des großen Einflusses des random state wird auch hier eine Zufallszahl zwischen 1 und 4e6 verwendet\n",
" randstate_data = np.random.randint(1,4000000)\n",
" X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.5, random_state=randstate_data) \n",
"\n",
" # Entscheidungsbaum erzeugen - auch hier Verwendung einer Zufallszahl\n",
" randstate_tree = np.random.randint(1,4000000)\n",
" clf = DecisionTreeClassifier(max_depth=6, random_state=randstate_tree) \n",
" clf.fit(X_train, y_train) # Entscheidungsbaum anhand der Trainingsdaten aufbauen\n",
" y_pred = clf.predict(X_test) # mit dem Entscheidungsbaum klassifizieren\n",
"\n",
" # erreichte Genauigkeit bestimmen\n",
" accuracy_vector_schlecht[i] =[metrics.accuracy_score(y_test, y_pred)]\n",
"\n",
" \n",
" "
]
},
{
"cell_type": "code",
"execution_count": 22,
"id": "9bce24c8-93f3-4400-ba19-3a87f14359d4",
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"plt.hist(accuracy_vector_schlecht,40,[0.5,1])\n",
"plt.hist(accuracy_vector_gut,40,[0.5,1])\n",
"plt.xlabel('Genauigkeit der Klassifikation')\n",
"plt.ylabel('Anzahl')\n",
"plt.grid('on')\n",
"plt.show()\n"
]
},
{
"cell_type": "markdown",
"id": "c30a352a-6cb2-4a34-9428-fa71d3b24408",
"metadata": {},
"source": [
"- - -\n",
"
Dieser Text steht unter der CC BY-SA 4.0-Lizenz. Der Name der Urheberin soll bei einer Weiterverwendung wie folgt genannt werden:
HTW Dresden, Fakultät Informatik/Mathematik "
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3 (ipykernel)",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.11.7"
}
},
"nbformat": 4,
"nbformat_minor": 5
}