1. Mining Weather Data Using R Data
Miner
Presented By,
Arbind Kumar (A13003)
Punit Kishore (A13011)

2. AGENDA
 Introduction
 Problem understanding
 Data understanding
 Rattle implementation
 Modeling
 Evaluation
 Conclusion

3. INTRODUCTION
 Weather is current & near future state of the
atmosphere given for a particular location.
 Weather is major factor in our daily decision
making.
 Affects us in the areas of our leisure activities,
transportation and communications, impacts on
working, shopping, our dress and fashion
coverings, and health/survival.
 In simple terms, we assess the condition, adapt
accordingly, and on our historical based
experience, we undertake decision making to try
and get to work or school safely without catching
a cold, having an accident or getting too wet.

4. PROBLEM UNDERSTANDING
 Here we are trying to predict the chance of rain
tomorrow.
 The above will cover:
 How does the weather will effect?
 What type of activities are effected by the weather?
 What is the commercial impact of the weather?
 How dependent are we on the weather?
 How at risk are we from the weather?

5. DATA UNDERSTANDING
 No of attributes: 24
 No of instances: 366
 Attribute information:
Variable Nam e
Meaning
Units
Date
Day of the month
Day
Location
Location of observations
Name
Min Temps
Minimum temperature in the 24 hours to 9am.
degrees Celsius
Max Temp
Maximum temperature in the 24 hours from 9am.
degrees Celsius
Rainfall
Precipitation (rainfall) in the 24 hours to 9am.
millimeters
Evaporation
Class A pan evaporation in the 24 hours to 9am
millimeters
Sunshine
Bright sunshine in the 24 hours to midnight
hours
WindGustDir
Direction of strongest gust in the 24 hours to midnight
16 compass points
WindGustSpeed
Speed of strongest w ind gust in the 24 hours to midnight
kilometers per hour
WindSpeed9am
Wind direction averaged over 10 minutes prior to 9 am
kilometers per hour
WindSpeed3pm
Wind direction averaged over 10 minutes prior to 3 pm
kilometers per hour
Humidity9am
Relative humidity at 9 am
percent
Humidity3pm
Relative humidity at 3 pm
percent
Pressure9am
Atmospheric pressure reduced to mean sea level at 9 am
hectopascals
Pressure3pm
Atmospheric pressure reduced to mean sea level at 3 pm
hectopascals
Cloud9am
Fraction of sky obscured by cloud at 9 am
eighths
Cloud3pm
Fraction of sky obscured by cloud at 3 pm
eighths
Temp9am
Temperature at 9 am
degrees Celsius
Temp3pm
Temperature at 3 pm
degrees Celsius
RainToday
Did it rain the day of the observation
Yes/No

6. RATTLE IMPLEMENTATION

7. MODELING

8.  Training instance is used to build a sequence
of evaluations that permits to determine the
correct category (prediction)
If pressure at 3pm >= 1012 hectopascals then
if the cloud at 3pm is < 7.5 then
it will not rain else
it may rain.
 Sequence of evaluations are represented as a
tree where leaves are labeled with the
category
 At each node, available attributes are
evaluated on the basis of separating the
classes of the training examples.

9. EVALUATION
Measure
Formula
Training
Accuracy, recognition rate
(TP+TN)/(P+N) 0.90
Error rate, misclassification rate
(FP+FN)/(P+N) 0.10
Sensitivity, true positive rate, recall
TP/P
0.63
Specificity, true negative rate
TN/N
0.95
Precision
TP/(TP+FP)
0.72
Test
0.81
0.19
0.60
0.87
0.58

10.  UNDERSTANDING DISTRIBUTION:

11. CONCLUSION
 This prediction may be helpful for the weather
forecasting centres as it predicts weather conditions
of a place.
 This will help for the following:
 Transportation-People tends to travel less in rain.
 Food
production-Many crops are sensitive to
temperatures & rain dependent.
 Health & Medical-Rain provides deinking water.
 Energy supply- Change in weather patterns can effect
energy production, especially in the newer
environmental supply systems like wind and solar.

12. THANK YOU