票据市场行情走势预测数据
收藏资源简介：
将票据市场需求信息采集、分类、归纳，提取有效数据，根据算法生成下一时刻数据，从而实现对市场行情进行预测算法规则：假设承兑行为K，K的初始价格A，承兑行修正参数m，直贴行修正参数n，市场波动系数β；上一个拟合价P;前6个时刻增幅的算术平均值α; (1)承兑行修正价格:B1;当K为国股时;B1=A;当K为五大时,B1=A+m;(2)直贴行修正价格:B2;当直贴类型为国贴时,B2=B1;偏离度修正价格:B3;计算:差值γ=B2-P;如果γ的绝对值-α的绝对值>β且如果γ>0则当前修正值B3=P+β;如果γ≤0则当前修正值B3=P-β;(4)计算拟合价C;当5分钟时间段内只有出或出、收都有价格,则C=出价格B3中的最大值;当间段内只有收的价格,则C=(收价格B3中的最小值+P)/2;(5)再次修正拟合价:D;计算:差值∆=C-P;如果差值∆的绝对值>β:如果∆>0,则:拟合价D=P+0.005;如果∆≤0,则:拟合价D=P-0.005;如果差值∆的绝对值≤β:拟合价D=C。
This dataset involves collecting, classifying, inducing and extracting valid data from demand information in the bill market, and generating data for the next moment through algorithms to realize market quotation prediction. The algorithm rules are as follows: Suppose the accepting bank is K, its initial price is A, the correction parameter of the accepting bank is m, the correction parameter of the direct discounting bank is n, the market fluctuation coefficient is β, the previous fitted price is P, and the arithmetic mean of the growth rates of the previous 6 moments is α. (1) Accepting bank corrected price B1: When K is a national state-owned shareholding bank, B1 = A; When K is one of the Big Five state-owned commercial banks, B1 = A + m; (2) Direct discounting bank corrected price B2: When the direct discounting type is national discount, B2 = B1; Deviation corrected price B3: Calculate the difference γ = B2 - P. If |γ| - |α| > β and γ > 0, then the current corrected value B3 = P + β; If γ ≤ 0, then the current corrected value B3 = P - β; (4) Calculate the fitted price C: When there are only bid prices or both bid and offer prices in the 5-minute time interval, C = the maximum value of the bid prices B3; When there are only offer prices in the time interval, C = (the minimum value of the offer prices B3 + P) / 2; (5) Revise the fitted price again to obtain D: Calculate the difference Δ = C - P. If |Δ| > β: if Δ > 0, then fitted price D = P + 0.005; if Δ ≤ 0, then fitted price D = P - 0.005; If |Δ| ≤ β, then fitted price D = C.
