Skip to content

pranjalsharma26/Playing-with-IRIS

Repository files navigation

Playing-with-IRIS Dataset On -

  1. Machine Learning Model
  2. Deep Learning Model
  3. Tensorflow

Before starting, do not forget to study the data which has 150 rows of 50 each types of flower and it is very easy to understand the data as well.


Types of IRIS flower

  1. Iris - Steosa
  2. iris - versicolor
  3. Iris - virginica

IRIS-Features-

  1. Petal length
  2. Sepal Length
  3. Petal width
  4. Sepal width

IRIS-Dataset on 3 different platforms -

  1. Machine Learning

1.a Here, the classification was done among comparing all the 4 features on two classifiers -

- KNeighborsClassifier

- DecisionTreeClassifier

1.b Calculation is done on these two models only and specifically all the comparisons between each of them are shown with defined accuracy in the notebook.

  1. Deep Learning

Simple DL model with 2 Dense layer is made along with Keras Classifier. Model turned out into 97% accuracy as well.

  1. Tensorflow

The result was around 100% in training whereas somehow dropped to 95.6% in validation set, which can be further improved.