review~

Naming alphabetically
graph pic include 


refernce me data set 
we reduce the dimension 
we predict  the missing value using the naive bayesin 
use digit for number rather than words


remove the "five" we will tell in next slides or so
try to succinct, concise language (motivation)

k=1,2,3....
we figured out k using "elbow method"  add elbow graph 


put that graph edge() if possible  

in k mean clustering graph occupation should be word
put a remark in the bottom saying that diff occupation are represented by dicrete value

k-nearest graph first part histogram 

knn ke two slides ko ek me merge karo 


last slides
remove fist point (make sure we have already talk about the ratio of test and training data set)

replace most of the term 

summary slides 
each point should be hardly one line 

ADD Thank you slides