Case 1:If we take observations at equal instants of time then
M
t1
= a.e
t1
M
t2
= a.e
t1+k
M
t3
= a.e
t1+2k
General term
M
tn
= a.e
t1+(n-1)k
i.e. the values of output Mforms a G.P. series of increasing order common ratio as
e
k
.Case 2:If we take observation at unequal timing interval in that caseT1 = t1 =>
M
t1
= a.e
t1
T2 = t1 + k1 =>
M
t2
= a.e
t1+k1
T3 = t2 + k2 = t1 + (k1 + k2) =>
M
t2
= a.e
t2+k2
= M
t2
=a.e
t1+k1+k2
General term Tn = T1+(k1+k2+k3+…+kn)
Tn = tn-1 + kn-1 = t1 +
(k1 + k2 + k3+…+kn
-1) =>
M
tn
= a.e
tn-1+kn-1
= M
tn
= a.e
(t1+k1+k2+k3+…+kn
-1)
= a.e
t1+Ktotal
i.e.
now any futuristic value say at instant tn is
M
tn
= a.e
t1
.e
Ktotal
(observed value)
Given
M
t
= a.e
t
, taking log on both sides we have,ln(M
t
) = ln(a) + ti.e. ln(M
tn
) = ln(a) + tnln(M
tn
) = ln(a) + t1+k1+k2+k3+…+kn
-1Thus we have obtained a log linear model for the abovefunction M
t
= a.e
t
using which we can calculate or predict thefuturistic values for increased ranges.Y = m.X + CIf we try to minimize the value of Ktotal we can do so by
making k1=k2=k3=…=kn
-1 which is same as Case 1.IV. PROJECTION OF SENSED INFORMATIONLet I= {i
1
,i
2
,…i
n
} be the set of sensed information. In theprocess of feature appropriate observation, forward selection ,backward elimination and decision based induction methodsare applied.
A. Forward selection based information sensing
Let I= {i
1
, i
2
,….,
i
n
}be the set of information estimates of various trends noted after observation in respective timinginstants Y = {y
1
,y
2
,…y
n
}. The accuracy measurement is to becalculated first based on comparison analysis. The minimumdeviation reflects high accuracy level of prediction and thatinformation will be selected. In this manner, { } , {bestinformation
},{first two}….will be s
elected.
B. Backward elimination based information sensing
Using backward elimination , in each stage each information iseliminated and thereby after the final screening stage theprojected set reveals the final optimum information space.
C. Cumulative frequency based information sensing
OBSERVATIONS INFORMATION INVOLVEDg
1
i
1
,i
3
,i
4
,i
6
g
2
i
3
,i
5
g
3
i
4
,i
5
,i
6
g
4
i
2
,i
3
,i
5
g
5
i
1
,i
2
g
6
i
1
,i
2
,i
3
,i
6
Table1 : Association of information against each observationFeatures InitialvalueCount Value (Value)
2
i
1
0.1 3 0.3 0.09i
2
0.2 3 0.6 0.36i
3
0.3 4 1.2 1.44i
4
0.4 2 0.8 0.64i
5
0.5 3 1.5 2.25i
6
0.6 3 1.8 3.24Table 2 : Determination of count and valueNow CF = ( x , y , z )where x = number of elements , y = linear sum of the elementsand z = sum of the square of the elements[3]V. BINARY TREE BASED GAIN CLASSIFIERIn this section information represents gain analysis. Asearch[4] can be formed based on the initial search term andits gradual sub term while the process of matching. Therebythe level is increased, in initial search term is the root and the
final term fully matching with the context of the users‟ desire
is a leaf node.In the above figure, G0 is the root that is initial search term. If a user wants to analyze further gain classification, thenidentify each search term as a binary code and by giving thecode number he can analyze the position of gain estimate inthe model . The concept of coding is as follows: