Trainer: Dr Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
https://en.akinator.com/
Trainer: Dr. Darshan Ingle
Impurity
Trainer: Dr. Darshan Ingle
Entropy
• Entropy is amount of information is needed to accurately describe the
some sample.
• So if sample is homogeneous, means all the element are similar than
Entropy is 0, else if sample is equally divided than entropy is
maximum 1.
Trainer: Dr. Darshan Ingle
Gini index / Gini impurity
• Gini index is measure of inequality in sample.
• It has value between 0 and 1. Gini index of value 0 means sample are
perfectly homogeneous and all element are similar, whereas, Gini
index of value 1 means maximal inequality among elements.
• It is sum of the square of the probabilities of each class.
Trainer: Dr. Darshan Ingle
So what is the importance of impurity measure in
decision tree??
• Impurity measures the homogeneity in the data sample. If the sample
is homogeneous then sample are from same class.
Trainer: Dr. Darshan Ingle
Lets start with weather data set, where target is to predict play
or not( Yes or No) based on weather condition.
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
So , lets focus on sub data on sunny outlook feature. we need to find the Gini index for temperature,
humidity and wind feature respectively.
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Now, Lets focus on sub data for overcast outlook feature.
Trainer: Dr. Darshan Ingle
Now,Lets focus on sub data for high and normal humidity feature.
Trainer: Dr. Darshan Ingle
Now, Lets focus on sub data for rainfall outlook feature. we need to find the Gini index for
temperature, humidity and wind feature respectively.
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Now, Lets focus on sub data strong and weak for wind rainfall feature.
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
Trainer: Dr. Darshan Ingle
