Task 1 – Decision Tree Classification using the Iris Dataset

Internship Project - Machine Learning (CodTech)

Hello! This repository contains my implementation of Task 1 from the Machine Learning Internship Program by CodTech. The task is to build, train, visualize, and evaluate a Decision Tree Classifier using the well-known Iris dataset.

This is the first of four tasks in the internship and focuses on classification using supervised learning.

Objective

To build a machine learning model that:

• Classifies different species of Iris flowers
• Uses four features: sepal length, sepal width, petal length, petal width
• Trains a Decision Tree Classifier using the scikit-learn library
• Evaluates its performance on unseen data
• Visualizes the decision tree structure for easy interpretability

Dataset Description

The dataset used is the Iris dataset, which is a classic in machine learning and pattern recognition.

• Features (Independent variables):

  • sepal_length (cm)
  • sepal_width (cm)
  • petal_length (cm)
  • petal_width (cm)

• Target (Dependent variable):

  • species (Setosa, Versicolor, Virginica)

In my version, the dataset is a .csv file uploaded and loaded via Google Colab.

📈 Step-by-Step Workflow

🔹 Step 1: Data Upload and Exploration

• Uploaded the dataset to Colab using files.upload()
• Loaded the CSV into a pandas DataFrame
• Explored the data using .head(), .info(), and .value_counts()

🔹 Step 2: Data Preprocessing

• Encoded the target column species into numeric format using LabelEncoder
• Split the dataset into features (X) and target (y)

🔹 Step 3: Train-Test Split

• Used train_test_split() from sklearn to divide data into 80% training and 20% testing sets

🔹 Step 4: Model Training

• Created and trained a DecisionTreeClassifier
• Fit the model on training data

🔹 Step 5: Prediction and Evaluation

• Predicted species for test data using .predict()
• Evaluated model using accuracy_score and classification_report

🔹 Step 6: Visualization

• Used plot_tree() to visualize the trained decision tree
• Displayed feature names and class labels for clarity

---

📊 Results

• Achieved high accuracy on the test data
• The decision tree structure shows clear and logical splits based on petal and sepal dimensions
• The model is interpretable, easy to explain, and performs well on small datasets like Iris

---

🛠️ Tools & Libraries Used

• Python 3
• Google Colab
• pandas
• scikit-learn
• matplotlib
• LabelEncoder

---

Output: