Weather

1
Weather

• By Shavar and Gerlof

2
Significance

• Predicting the weather creates a huge benefit for
  a large amount of people.
• Harsh weather can cause a disaster in the
  agrarian sector, in case of outdoor events or to
  civilization in general, like we have seen in
  Asia.
• Therefore its crucial to be able to predict
  weather accurately

3
Problem

• As youve seen it can lead to awkward situations
  if the weather isnt known before it takes place.
• In our data set the main goal is to predict if
• outdoor sports games would continue

4
Attributes of our data

• Outlook
• Temperature
• Humidity
• Windy
• Play (whether the sports team plays or not)

Sunny Overcast rainy
5
Interesting Patterns (by analyzing attributes)

• Most of the time the temperature is below 72.4
  degrees Fahrenheit
• Almost two times more games are being played then
  being cancelled.
• A lot of plays still go on even when its
  raining!!
• And even windiness cant stop players from
  playing their outdoor sports

6
Classification Acc.

• We decided to use a decision tree in order to
  predict if the plays would continue or not
• The reliability of this was about 50, just as it
  is on TV as we all know.

7
Interesting Patterns(by analyzing classifiers)

• The prediction became more effective by using a
  more advanced tree.
• Because of the relatively small amount of
  attributes, there couldnt be found any more
  patterns.

8
In Practice

• Predicting the weather is used a lot in public
  already, as we all know on the television, and by
  analyzing data we should be able to refine our
  methods.