Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Abstract––Partitioning the universe of discourse and different lengths of intervals may affect the accuracy of
determining effective intervals are critical for forecasting forecasting. Recently, Ref.[17] used particle swarm
in fuzzy time series. Equal length intervals used in most optimization method to search for a suitable partition of
existing literatures are convenient but subjective to universe. Additionally, Ref. [18] proposed a new method
partition the universe of discourse. In this paper, a new to forecast enrolments based on automatic clustering
forecasting model based on two computational methods, techniques and fuzzy logical relationships. In this paper,
the fuzzy logical relationship groups and automatic we proposed a new forecasting model combining the
clustering technique is presented for forecasting fuzzy time series and automatic clustering technique. The
Gasonline Price. Firstly, we use the automatic clustering method is different from the approach [3] and [17]in the
algorithm to divide the historical data into clusters and way where the fuzzy relationship groups are created.
adjust them into intervals with different length. Then, Based on the model proposed in [20], we have developed
based on the new obtained intervals, we fuzzify all the a new fuzzy time series model by combining the
historical data into fuzzy sets and calculate the automatic clustering technique and fuzzy time series with
forecasted output by the proposed method. To illustrate the aim to increase the accuracy of the forecasting model.
the forecasting process, a numerical data sets of
Gasonline Prices in Viet Nam is utilized to calculate by The rest of this paper is organized as follows. In Section
step. The results show that the proposed model gets II, we provide a brief review of fuzzy time series and it‟s
forecasting accuracy for a higher high – order compare model. In Section III, we present a model for forecasting
with first – order fuzzy time series model. the Gasonline price in VietNam based on the automatic
clustering technique and fuzzy time series. Then, the
Index Terms––Forecasting, FTS, fuzzy logical experimental results are shown and analyzed in Section
relationships, automatic clustering, Gasonline prices IV. Finally, Conclusions are given in Section V.
1) → F(t). According to[3]suggested that when the The evaluated fuzzy time-series model which applied
maximum degree of membership of F(t) belongs toAi , simplified arithmetic operations in forecasting
F(t) is considered Aj . Hence, the relationship between algorithms rather than the complicated max-min
F(t) and F(t -1) is denoted by fuzzy logical relationship composition operations was proposed by Chen [3]. In
addition, Chen‟s method can generate more precise
A i → Aj where Ai and Aj refer to the current state or the
forecasting results than those of Song and Chissom[1].
left - hand side and the next state or the right-hand side
Chen‟s method consists of the following major steps:
of fuzzy time series.
Step 1: Define the universe of discourse U.
Definition 3: 𝛾- order fuzzy logical relationships
Step 2: Divide U into several equal-length intervals.
Let 𝐹(𝑡) be a fuzzy time series. If 𝐹(𝑡) is caused by Step 3: Define the fuzzy sets on U and fuzzify the
𝐹(𝑡 − 1), 𝐹(𝑡 − 2), … , 𝐹(𝑡 − 𝛾 + 1) 𝐹(𝑡 − 𝑚) then this historical data.
fuzzy relationship is represented by by 𝐹(𝑡 − Step 4: Create the fuzzy logical relationships based on
𝛾), … , 𝐹(𝑡 − 2), 𝐹(𝑡 − 1) → 𝐹(𝑡) and is called an 𝑚 - the historical data.
order fuzzy time series. Step 5: Classify the derived fuzzy logical relationships
Definition 4: Fuzzy relationship group (FRG) into groups.
Fuzzy logical relationships, which have the same left- Step 6: Forecast and defuzzify output value
hand sides, can be grouped together into fuzzy logical
relationship groups. Suppose there are relationships such
as follows:
A PROPOSED FORECASTING MODEL
𝐴𝑖 → 𝐴𝑘1 , 𝐴𝑖 → 𝐴𝑘2 , ……. BASED THE FUZZY TIME SERIES AND
AUTOMATIC CLUSTERING
In previous study was proposed by Chen [3], the ALGORITHM
repeated fuzzy relations were simply ignored when fuzzy In this section, an improved hybrid model for forecasting
relationships were established. So, these fuzzy logical the Gasonline Pricebased on clustering technique and
relationships can be grouped into the same FRG as: FTS. Firstly, we apply clustering technique to classify
𝐴𝑖 → 𝐴𝑘1 , 𝐴𝑘2 … the collected data into clusters and adjust these clusters
into contiguous intervals, based on the new intervals
Fuzzy time series forecasting model defined, we fuzzify on the historical data, determine
Fuzzy theory [19] was developed to deal with problems fuzzy logical relationships and create fuzzy logical
involving linguistic terms. Song and Chissom[1], [2] the relationship groups and calculate forecasted value
application of this theory to define fuzzy time-series according to the forecasting model of Chen [3].
model. They applied the model to forecast the
enrollments of the University of Alabama. Further, a To verify the effectiveness of the proposed model, all
first-order time-invariant fuzzy time series model was historical data sets of Gasonline Price of E5 RON 92of
proposed by Song and Chissom to forecast the Viet Nam are shown in Fig. 1, which are used to
enrollments and a step-by-step procedure is presented to illustrate for forecasting process. The step-wise
develop and utilize the time-variant model [2]. procedure of the proposed model is detailed as follows:
Table 3:The complete result of the first- order fuzzy logical relationships.
No linguistic value First –order fuzzy relations No linguistic value First –order fuzzy relations
1 A1 A1 -> A3 11 A7 A7 -> A7
2 A3 A3 -> A3 12 A6 A7 -> A6
3 A3 A3 -> A2 13 A6 A6 -> A6
4 A2 A2 -> A2 14 A6 A6 -> A6
5 A2 A2 -> A2 15 A6 A6 -> A6
6 A2 A2 -> A4 16 A6 A6 -> A6
7 A4 A4 -> A5 17 A6 A6 -> A6
8 A4 A5 -> A7 18 A7 A6 -> A7
9 A5 A13 -> A12 19 A8 A7 -> A8
10 A7 A12 -> A9 20 A9 A8-> A9
Table 4: The complete result of the first - order fuzzy relationship groups.
No linguistic value First –order fuzzy relations No linguistic value First –order fuzzy relations
1 A1 A1 -> A3 11 A7 A7 -> A7
2 A3 A3 -> A3 12 A6 A7 -> A6
3 A3 A3 -> A3,A2 13 A6 A6 -> A6
4 A2 A2 -> A2 14 A6 A6 -> A6, A6
5 A2 A2 -> A2, A2 15 A6 A6 -> A6, A6, A6
6 A2 A2 -> A2, A2 16 A6 A6 -> A6, A6, A6,A6
7 A4 A2 -> A2, A2, A4 17 A6 A6 -> A6, A6, A6, A6,A6
8 A4 A4 -> A4 18 A7 A6 -> A6, A6, A6, A6,A6, A7
9 A5 A4 -> A5 19 A8 A7 -> A7, A6, A8
10 A7 A5 -> A7 20 A9 A8-> A9
Table 5: The complete forecasted outputs for gasonline price of Viet Nam based on the first– order FTS model.
Date Actual data linguistic value Forecasted value Date Actual data linguistic value Forecasted value
4/1/2018 18240 A1 ---------- 7/6/2018 19940 A7 19930
19/1/2018 18670 A3 18652.5 22/6/2018 19610 A6 19743.33
3/2/2018 18670 A3 18652.5 7/7/2018 19610 A6 19650
21/2/2018 18340 A2 18476.5 23/7/2018 19610 A6 19650
8/3/2018 18340 A2 18388.5 28/7/1998 19610 A6 19650
23/3/2018 18340 A2 18388.5 7/8/2018 19610 A6 19650
7/4/2018 18930 A4 18690.5 22/8/2018 19610 A6 19650
23/4/2018 18930 A4 18992.5 6/9/2018 19910 A7 19730
8/5/2018 19440 A5 19234.17 21/9/2018 20230 A8 20034.92
23/5/2018 19940 A7 19930 06/10/2018 20906 A9 20780.88
Step 9: Calculate and defuzzify the forecasted output Based on the forecasted rules above and based on Table
values 4, we complete forecasted results of Gasonline price in
In this step, to obtain the forecasted results, we use Viet Nam based on first - order FTS model with nine
defuzzification technique [20] to calculate the forecasted intervals are listed in Table 5.
values for all date of Gasonline price in training phase.
The defuzzification rules are presented for date t The performance of proposed model can be assessed by
(04/01/2018 ≤ t ≤ 06/10/2018) as follows: comparing the difference between the forecasted values
Rule 1: If the fuzzified data of Gasonline price on date and the actual values. The mean square error (MSE) is
t-1 is Aj and there is only one fuzzy logical relationship employed as an evaluation criterion to represent the
in the fuzzy logical relationship group whose current forecasted accuracy. The MSE value is computed as
state is Aj, shown as follows: A j (t − 1) → Ak (t) , then follows:
1 𝑛
the forecasted Gasonline price of datet is mk, where mk 𝑀𝑆𝐸 = (𝐹 − 𝑅𝑡 )2 (12)
𝑛 𝑡=γ 𝑡
is the midpoint of the interval uk and the maximum Where, 𝑅𝑡 denotes actual value at year t, 𝐹𝑡 is forecasted
membership value of the fuzzy set Ak occurs at the value at year t, n is number of the forecasted data, γ is
interval uk. order of the fuzzy logical relationships
Rule 2: If the fuzzified Gasonline price of datett -1 is Aj
and there are the following fuzzy logical relationship COMPUTATIONAL RESULTS
group whose current state is Aj, shown as follows: In this paper, we apply the proposed model to forecast the
Aj t − 1 → Ai1 t1 , A i2 t2 , Aip tk gasonline price of Viet Nam with the whole historical data
then the forecasted Gasonline price of datett is calculated the period from 4/1/2018 to 6/10/2018 is shown in Fig.1.
1∗m i1 +2∗m i2 +3∗m i3 +⋯+p∗m ip We implement the proposed model under different
as follows: forecasted = ;
1+2+⋯+p number of orders and kept number of intervals of 9. The
where 𝑚𝑖1 , 𝑚𝑖2 , 𝑚𝑖𝑘 are the middle values of the forecasted accuracy of the proposed method is estimated
intervals ui1 , ui2 and uip respectively, and the by using the MSE function (12).The forecasted results of
maximum membership values of Ai1 , Ai2 , . . . , Aip occur proposed model under number of interval as 9 and various
at intervals ui1 , ui2 , . . . , uip , respectively. order of fuzzy relationship are listed in Table 6.
Table 6: The completed forecasting results for gasonline price data of Viet Nam under deferent orders of FTS.
Date Actual data Forecasted value for each order of fuzzy logicalrelationship
2nd-order 3rd-order 4th-order 5th-order 6th-order 7th-order
4/1/2018 18240 ----- ----- ----- ----- ----- -----
19/1/2018 18670 ----- ----- ------ ------ ----- ------
3/2/2018 18670 18652.5 ----- ----- ------ ----- ------
21/2/2018 18340 18388.5 18388.5 ---- ------ ----- ------
8/3/2018 18340 18388.5 18388.5 18388.5 ----- ----- -----
23/3/2018 18340 18388.5 18388.5 18388.5 18388.5 ----- -----
7/4/2018 18930 18791.17 18992.5 18992.5 18992.5 18992.5 -----
23/4/2018 18930 18992.5 18992.5 18992.5 18992.5 18992.5 18992.5
8/5/2018 19440 19355 19355 19355 19355 19355 19355
23/5/2018 19940 19930 19930 19930 19930 19930 19930
7/6/2018 19940 19930 19930 19930 19930 19930 19930
22/6/2018 19610 19650 19650 19650 19650 19650 19650
7/7/2018 19610 19650 19650 19650 19650 19650 19650
23/7/2018 19610 19650 19650 19650 19650 19650 19650
28/7/1998 19610 19650 19650 19650 19650 19650 19650
7/8/2018 19610 19650 19650 19650 19650 19650 19650
22/8/2018 19610 19650 19650 19650 19650 19650 19650
6/9/2018 19910 19743.33 19762 19790 19836.67 19930 19930
21/9/2018 20230 20326.5 20326.5 20326.5 20326.5 20326.5 20326.5
06/10/2018 20906 20780.88 20780.88 20780.88 20780.88 20780.88 20780.88
MSE 5573 4633 4307 3836 3536 3561
21000
Actaul Data of Gasonline price
20500
20000
17000
18240
18670
18670
18340
18340
18340
18930
18930
19440
19940
19940
19610
19610
19610
19610
19610
19610
19910
20230
20906
Date
Fig. 2: A comparison of the MSE values for 9 intervals with different high-order fuzzy relationships.
[13] Liu, H.T., "An Improved fuzzy Time Series [20] Nghiem Van Tinh & Nguyen Cong Dieu,
Forecasting Method using Trapezoidal Fuzzy improving the forecasted accuracy of model based
Numbers," Fuzzy Optimization Decision Making, on fuzzy time series and k-means clustering,
Vol. 6, pp. 63–80, 2007. journal of science and technology: issue on
[14] Lee, L.-W., Wang, L.-H., & Chen, S.-M. information and communications technology,
Temperature prediction and TAIFEX forecasting No.2, 51-60, 2017
based on fuzzy logical relationships and genetic
algorithms. Expert Systems with Applications, 33, Biography of Author(s)
539–550, 2007. Nghiem Van Tinh received the B.S. degree in applied
[15] Jilani, T.A., Burney, S.M.A. A refined fuzzy time mathematics and information from HaNoi University of
series model for stock market forecasting. Physica Science and Technology (HUST), VietNam, in 2002.
A 387, 2857–2862. 2008. The M.S. degree in Information of Technology from
[16] Wang, N.-Y, & Chen, S.-M. Temperature Thai Nguyen University of Information and
prediction and TAIFEX forecasting based on Communication Technology, in 2007. He is currently a
automatic clustering techniques and two-factors Ph.D. student at Institute of Information Technology,
high-order fuzzy time series. Expert Systems with Vietnam Academy of Science. He is currently work at
Applications, 36, 2143–2154, 2009. Faculty of Electronics Engineering, ThaiNguyen
[17] Kuo, I. H., Horng, S.-J., Kao, T.-W., Lin, T.-L., University of Technology, ThaiNguyen, VietNam.
Lee, C.-L., & Pan. An improved method for
forecasting enrollments based on fuzzy time series Nguyen Tien Duy received the Ph.D degree in applied
and particle swarm optimization. Expert Systems mathematics from Institute of Information Technology
with applications, 36, 6108–6117, 2009a. (IOIT), Vietnamese Academy of Science and
[18] S.-M. Chen, K. Tanuwijaya, Fuzzy forecasting Technology, in 2016. The M.S. degree in Information of
based on high-order fuzzy logical relationships Technology from Thai Nguyen University of
and automatic clustering techniques, Expert Information and Communication Technology, in 2007.
Systems with Applications 38 ,15425–15437, The B.S degree in Information Technology, Hanoi
2011. University of Science and Technology, Hanoi, Vietnam,
[19] Zadeh, L. A. . Fuzzy sets. Information and Control, 2001. He is working at Faculty of Electronics,
8, 338–353, 1965. ThaiNguyen University of Technology, ThaiNguyen,
VietNam.