forked from acredjb/FBP
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Advantages
40 lines (30 loc) · 2.87 KB
/
Advantages
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
1. The accuracy rate is high, reaching 80%.
2. Predition time is early, can be made 1 days ahead of FootBall Game schedule.
3. Not affected by personal emotions, the predition result is entirely exported by the program.
4. Do not be affected by trader's hand, ignore the change of the plate.
5. The prediction model is a process of "iteration and correction of historical odds combination", which conforms to
the theory of back propagation algorithm in machine learning and predicts the result of large data odds combination.
6. Updating forecasting model data constantly, with the gradual increase of historical data,
the accuracy will continue to improve,and stable in high accuracy.
7. Systematically describe predictive rules, process historical data, observe how predictions have behaved in the past,
and then rationally improve these predictive rules according to specific results.
8. It will also incorporate new rules and predictive rules that may be overlooked by the program
(e.g. program 1.0 only considers the odds of major spinach companies, program 2.0 adds dimension to the league,
and program 3.0 adds periodic data as a new predictive rule).
At this point, the prediction ability of the program is much improved, more precise, faster, and free from emotional interference.
9. It is not blindly following the suggestion of procedure, but at the same time of procedure prediction,
I also analyze (combining the basic aspects, especially the historical confrontation), and then compare the two.
When the decision of the procedure is different from mine, I will check the reason.
10.author email:[email protected]
Welcome to send email to me.
1、准确率高,达到80%左右
2、时间早,可以提前1天做出程序预测
3、不受个人情绪影响,完全由程序输出看好的比赛
4、不受操盘手影响,无视盘口的变化
5、预测模型是一个“历史赔率组合不断迭代修正”的过程,符合机器学习中的反向传播算法理论,以大数据赔率组合情况产生的结果做预测
6、不断更新预测模型的数据,随着历史数据的逐渐增加,准确率会不断提高,并稳定在高准确率
7、系统化的方式描述预测规则,处理历史数据,观察预测在过去的表现如何,进而根据具体结果合理的改进这些预测规则
8、还会结合新的规则、程序目前可能忽略的预测规则(如程序1.0版本只考虑了各大菠菜公司的赔率,程序2.0就增加了联赛的维度,程序3.0又增加了阶段性数据作为新的预测规则),
此时的程序预测能力就提高了很多,更精确、更快、更不受情绪干扰
9、并不是盲从程序的建议,而在程序预测的同时,本人也进行分析(结合基本面尤其是历史交锋),然后对二者进行对比。当程序的决策与我不同,我会检查原因。
10、可以通过email联系作者。