1: \begin{abstract}

2: With the rapid growth of data, how to extract effective information from data is one of the most fundamental problems.

3: In this paper, based on Tikhonov regularization, we propose an effective method for reconstructing the function and its derivative from scattered data with random noise.

4: Since the noise level is not assumed small, we will use the amount of data for reducing the random error, and use a relatively small number of knots for interpolation.

5: An indicator function for our algorithm is constructed.

6: It indicates where the numerical results are good or may not be good.

7: The corresponding error estimates are obtained.

8: We show how to choose the number of interpolation knots in the reconstruction process for balancing the random errors and interpolation errors.

9: Numerical examples show the effectiveness and rapidity of our method.

10: It should be remarked that the algorithm in this paper can be used for on-line data.

11: % We develop a big data processing technique that reconstructs a function and its derivative from sample values with random noise.

12: % The solution space of function reconstruction is a finite dimensional vector space whose dimension is relatively small, which reduces computational burden.

13: % With a given dataset, we design an indicator function which shows the effective regions where reconstruction results are reliable.

14: % We also give asymptotic convergence rates, and provide numerical examples.

15: \end{abstract}

16: