Keras-Neural-Network-Analysis-Iris-Dataset

Classes and Functions used: Numpy, Pandas, KerasClassifier, Cross validation score, K fold validation and Label Encoder

This is my first attempt in defining a keras classifier for multi-class classification, and is done by the Iris Dataset(Most popular for Novie Machine Learning Enthusiasts) on the UCI Machine Learning Repository. Link for data set: http://archive.ics.uci.edu/ml/datasets/Iris

Algorithm:

1. Fixed the random seed variable at 10.
2. Loaded the dataset from iris.csv and split it into input variables and output variables.
3. Did the one hot encoding or created dummy variables from a categorical variable using scikit-learn class LabelEncoder and then we convert the vector of integers to a one hot encoding using the Keras function to_categorical().
4. Defined the Neural Network Model(Keras Classifier) {4 input variables -> [4 nodes in hidden layer] -> 3 output variables}
5. Evaluated the model using K-Fold Cross Validation and also computed the accuracy of the classifier.

Under update! please be patient