There are many goals of data science. Some of the important goals are the following.

• Insight
• Prediction

Insight provides a better appreciation of the data for decision-making purposes.

In the Money Ball story, the insight from the data was which parameters to take as more important than other parameters when making decisions (e.g., which ballplayer to sign to a contract and for how much).

Prediction is to claim or decide that something will happen in the future or as a result of further investigation.

• A forecast is a prediction of what will happen in the future. As former major league baseball catcher Yogi Berra is reported to have said, "Prediction is hard, especially in the future.". Everything is easier with 20/20 hindsight.

In general, an interpolation is more accurate than an extrapolation.

An interpolation is determining a value within the region of available data. For example, determining y given an x that is within the available values of x.

A Bezier curve is specified by several points and successive and recursive interpolation is done until the curve is sufficiently "curved" as seen by the human eye.

Note: In graphics systems such as SVG, a circle is approximated by four Bezier curves. The human eye can not distinguish between the circle and the Bezier approximation.

An extrapolation is determining a value the goes beyond the region of available data. For example, given the past history, predict the future via extrapolation. There tends to be more error associated with extrapolation than with interpolation.

A catastrophe is a sudden change that is not foreseen.