Reference no: EM133339225
In this problem use the seaborn Python package to create a scatter plot matrix of the Iris data set.
1. Describe the diagonal curves.,
2. Does any feature separate the three classes? If so, which feature? If not, which is the closest?,
3. Is there any combination of features that can separate the three classes? If so, which features? If not, which are the closest?,
In this problem, implement code to use the Mahalanobis Distance on the Iris dataset.
1. Using all four features, calculate the 50 distances for the Setosa class.
2. Calculate the range (min and max) of the 50 Mahalanobis distance values. What is the delta between the min and the max?
3. If you remove the observation with the largest distance, do the remaining distances appear to be closer together? After removing the observation with the largest distance, recalculate the ranges with the updated mean and covariance. Does removing the observation with the largest Mahalanobis distance provide a smaller range between the min and max?
4. Given your findings, describe what was accomplished following the method above this data point from the dataset.
"In this problem we would like to familiarize you with a basic example of a goal based agent. A game like rock paper scissors is typically played between two players and has a clearly defined set of rules for a win, draw, and loss. Since it is impossible (acknowledging that studies have been preformed to use behavior to attempt a prediction - but we will ignore that for the purpose of this exercise) to predict an opponent's move, we will assume that the user will input a value, the computer will see this input then choose from a set of rules the correct response play, acting as a simple goal based agent.
Instructions on how to play: wikihow Play-Rock,-Paper,-Scissors
For this problem, you will fill in the rules (in if-then-else format) to the function in the cell below. The proceeding cell should then be run to play a game with the computer in which the computer agent should win every time.
Attachment:- problems.rar