farah-dhiaf / ML-IRIS-classification

Iris classification exercise

ML-IRIS-classification

Nama : Farah Dhia Fadhila
NRP : 5025211030
Kelas : Pembelajaran Mesin C

Classification Iris flower dataset into 3 classes (Iris Setosa, Iris virginica, and Iris versicolor) using classification method and artificial neural network. There are 2 scenarios, with normalization using MinMax scalling and without normalization.

1. Decision Tree
   • Criterion gini
   • Criterion entropy
2. KNN
   • k=3
   • k=5
   • k=7
3. Naive Bayes
4. SVM
   • kernel linear
   • kernel polynomial
   • kernel rbf
   • kernel sigmoid
5. ANN
   • 2 hidden layers (6 neurons for each layer, activation function ReLU)
   • 1 output layers (3 neuron, activation function softmax)
   • Compiling (optimizer="adam", loss="categorical_crossentropy", metrics=['accuracy']), fitting (batch_size=32, epochs=100)