NEWAY

愿识乾坤大，仍怜草木青

0%

ENGG5189 fuzzy query

Posted on Edited on In Valine:

NAME: Li Wei

SID: 1155062148

Criteria design

In this section, I will introduce and justify my criteria and state my assumptions I make when designing this kind of criteria. Typically, fuzzy criteria will be highlighted.

  • The stock with highest PE in same type will be excluded. This criteria is inspired by the introduction of PE in the specification:

    PE(Price-Earning Ratio) = Market price / EPS(earning per share). Usually lower is better, but it depends on the PE ratio of the similar stocks. (However, if some types contain only one stock in the database, then this special stock won’t be excluded.)

  • high buying level. High buying level means more stable and has less risk. The following is the corresponding fuzzy membership function.

    Buying level 1 2 3 4 5
    $\mu_{high_buy_lv}$ 0.1 0.3 0.5 0.7 0.9

  • high difference. High difference level means that the stock may have more profit but also more risk. The following is the corresponding fuzzy membership function. However, the difference between ‘too high’ and ‘very high’ is not huge. Therefore the fuzzy memebership function is not linear.

    Difference 0 25 50 75 100 125 150 175 200 225 250
    $\mu_{high_diff}$ 0.0 0.2 0.4 0.6 0.75 0.8 0.85 0.9 0.95 0.98 1.0

Query

This section is to introduce the query we design:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
STOCK_CLASS(NAME) (or ((and (= Type SEMI) (< PE 31))
                       (and (= Type WWW) (< PE 66.4))
                       (= Type TECH)
                       (= Type FOOD)
                       (= Type GAME)
                   )
                   (and 
                   	   (High_BUY_LV
                       (1 2 3 4 5)
                       (0.1 0.3 0.5 0.7 0.9)1)
                   	   (High_DIFF_LV
                       (0 25 50 75 100 125 150 175 200 225 250)
                       (0.0	0.2	0.4 0.6 0.75 0.8 0.85 0.9 0.95 0.98 1.0)1)
                   )

Calculation

Non-fuzzy criteria

The stock with highest PE in same type will be excluded.

NAME PRICE ORIGI N TYPE PE PE_LV BUY_LV DIFF DIFF_LV
AAPL 167 US TECH 18 1 4 33 2
TXN 110 US SEMI 31 2 4 47 2
FB 187 US WWW 35 2 5 43 2
MU 44 US SEMI 7 1 5 80 3
BABA 205 CHN WWW 55 4 5 101 4
KM 120 CHN FOOD 39 3 4 220 5
SAMSUNG 2325 KOR SEMI 9.53 1 4 25 1
NINTENDO 440 JPN GAME 50 3 4 200 5
TENCENT 59 CHN WWW 58 4 4 230 5
PYPL 85 US WWW 66.4 5 3 90 4
Fuzzy criteria

high buying level and high difference level.

Name Buying level Difference level Membership(min)
AAPL $\mu_{high_buy_lv}(4) = 0.7$ $\mu_{high_diff}(33)=0.2$ 0.2
FB $\mu_{high_buy_lv}(5) = 0.9$ $\mu_{high_diff}(43)=0.2$ 0.2
MU $\mu_{high_buy_lv}(5) = 0.9$ $\mu_{high_diff}(80)=0.6$ 0.6
BABA $\mu_{high_buy_lv}(5) = 0.9$ $\mu_{high_diff}(101)=0.75$ 0.75
KM $\mu_{high_buy_lv}(4) = 0.7$ $\mu_{high_diff}(220)=0.95$ 0.7
SAMSUNG $\mu_{high_buy_lv}(4) = 0.7$ $\mu_{high_diff}(25)=0.2$ 0.2
MINTENDO $\mu_{high_buy_lv}(4) = 0.7$ $\mu_{high_diff}(200)=0.95$ 0.7
TENCENT $\mu_{high_buy_lv}(4) = 0.7$ $\mu_{high_diff}(230)=0.9$ 0.7

Conclusion

Therefore, the system will select BABA to buy from our calculation.