The curve which is obtained by plotting a set of data points is called Curve Fitting.
The curves like circular, elliptical, parabolic, and hyperbolic. Sine, cosine, etc. are also obtained by fitting data points.

In the given figure, a curve is drawn through several points.

A. W = 495(4)t
B. W = 360t2 + 1125t + 495
C. W = 1485t + 495
D. W = 360t2 + 495 Correct Answer: B
Step 1: Plot the points and join with a smooth curve.
Step 2: The graph appears to be curved, not linear.
Step 3: As the ratios of consecutive number of workers is not same,
an exponential model does not fit the data.
step 4: W = at2 + bt + c [Quadratic model.]
Step 5: Consider the three data pairs from the table (0, 495),
(1, 1980), and (2, 4185) and substitute in the model.
Step 6: The equations obtained are
c = 495 ------------ (1)
a + b + c = 1980 ----------- (2)
4a + 2b + c = 4185 --------- (3)
Step 7: a = 360, b = 1125, and c = 495 [Solve the simultaneous equations.]
Step 8: So, the number of workers in the industry modeled by W = 360t2 + 1125t + 495.
Q1: Which of the following is the primary goal of curve fitting?
Q2: What is a potential drawback of overfitting data in curve fitting?
Q: How do I choose the best curve for my data?
A: Examine the scatter plot of the data to identify trends. Consider the underlying physics or principles of the system being modeled.
Q: What is overfitting?
A: Overfitting occurs when the curve fits the data too closely, capturing noise and random variations instead of the underlying pattern. This leads to poor predictions for new data.