Year =1:10;

p=[

46.2 1

32.6 2

26.7 3

23.0 4

20.0 5

18.9 6

17.5 7

16.3 8

15.2357 9

14.4650 10

]';

t=[

32.6

26.7

23.0

20.0

18.9

17.5

16.3

15.2357

14.4650

13.8732

]';

% to normalize the original data.

[pn，meanp，stdp，tn，meaned，stdt]=prestd(p，t)；

% to establish the corresponding BP network

net = newff(minmax(pn)，[7， 1]，{'tansig' 'purelin' }，' train gdx ')；

% training network

net . train param . epochs = 2000；

net . train param . goal = 0.000 1；

net = train(net，pn，TN)；

% to simulate the trained network.

an=sim(net，pn)；

a=poststd(an，meant，stdt)；

% draw a simulated image

Plot (year, t,' b', year, a,' r');

Title ("Analog Image")

p _ new =[ 13.5 13]'；

pn_new=trastd(p_new，meanp，stdp)；

an_new=sim(net，pn _ new)；

a=poststd(an_new，meant，stdt)

This is a neural network. You can try to predict the data.