Using gradient descent (a common machine learning technique) to train and perform prediction on the iris dataset.
Based on the training dataset, we aim at doing prediction on the test dataset.
For the data we will be dealing with, there will be 3 labels: ’setosa', 'versicolor' and ‘virginica'.
We will have to implement an algorithm to predict the correct label based on the given data attributes, namely, the values of sepal_length, sepal_width, petal_length, and petal_width. This is a classification problem. To solve the problem, you will need to implement a particular model. We will use softmax regression (which is similar to logistic regression, but generalises to multi-classes) and do gradient descent.
Step 1: Getting the desired data from the raw data
Step 2: Training the model based on the training data we got from Step 1
Step 3: Predicting and testing accuracy based on the trained model from Step 2
Step 4: Using the sklearn library to evaluate the accuracy at which sklearn performs.