This dataset was obtained from the UCI machine learning repository
(http://www.ics.uci.edu/~mlearn/MLSummary.html).
The dataset at UCI had originally 9 attributes (the first being the goal
variable, the mpg attribute). The last attribute (the car model) is unique for
all instances so I've REMOVED IT!
I've also moved the first attribute (the goal variable) to the end, so now is
the last attribute, just for compatibility with my dataset format.
All other things are as in UCI.
In the repository they referred an original dataset were apparently there were
8 more instances, but they were removed because they had the "mpg" attribute
(the goal) with unknown value.
The attributes brief description goes as follows :
1. cylinders: multi-valued discrete # the number of cylinders
2. displacement: continuous # ?
3. horsepower: continuous # the power of the engine
4. weight: continuous # obvious meaning
5. acceleration: continuous # obvious meaning
6. model year: multi-valued discrete # obvious meaning
7. origin: multi-valued discrete # obvious meaning
8. mpg: continuous # I think is the fuel consumption (miles per
gallon)