Iris_ANN

This notebook builds an Artificial Neural Network (ANN) using TensorFlow/ to classify Iris flower species based on sepal and petal measurements.

Key Steps:

Data Loading & Exploration:

Loads the Iris dataset (150 samples, 4 features).

Drops the Id column, checks for missing values/duplicates.

Visualizes feature distributions and pairwise relationships.

Preprocessing:

Encodes the target (Species) using LabelEncoder and converts to one-hot format.

Standardizes features using StandardScaler.

Splits data into training and test sets (80/20 split).

Model Architecture:

Input layer: 4 features → 16 neurons (ReLU).

Hidden layer: 8 neurons (ReLU).

Output layer: 3 neurons (Softmax for multi-class classification).

Training:

Compiled with Adam optimizer and categorical cross entropy loss.

Trained for 50 epochs with batch size 8 and 20% validation split.

Evaluation:

Achieves ~97% test accuracy.

Confusion matrix and classification report show perfect precision/recall for most classes.