Table 22 Variables and the corresponding definition
From: A hybrid model for stock price prediction based on multi-view heterogeneous data
Variable | Definition |
|---|---|
\(f_{t,m}\) | The occurrence frequency of \(word_m\) in \(news_t\) |
\(F_m\) | The occurrence frequency of \(word_m\) in the news corpus |
A | The number of times that word w and category c co-occur |
B | The number of times that word w occurs without category c |
C | The number of times that category c occurs without word w |
D | The number of times that neither category c nor word w occurs |
P(c) | The frequency of category \(c \in \{-1,1\}\) in news corpus |
\(\xi _i\) | A slack variable |
C | A penalty term controlling the cost to misclassification of samples |
\(\alpha _{i}\) | A Lagrangian multiplier corresponding to sample \(x_i\) |
\(k(x_{i},x_{j})\) | The kernel function |
\(\gamma\) | A Gaussian kernel parameter |
\(\beta _{m}\) | The weight of kernel \(k_{m}(x_{i},x_{j})\) |
\(o_{t}\) | The first order differencing of a data series \(z_t\) |
p | The autoregression order |
d | The differencing order |
q | The moving average order |
\(\phi _{i}\) | The i-th autoregression parameter |
\(\theta _{j}\) | The j-th moving average parameter |
\(\epsilon _{t}\) | The error term at time t |
\(r_t\) | Stock daily return on \(day_t\) |
\(tv_t\) | Trading volume on \(day_t\) |
\(tr_t\) | Turnover rate on \(day_t\) |
\(mc_t\) | Market cap on \(day_t\) |
\(md_t=(r_t, tv_t, tr_t, mc_t)\) | The market data including four market variable on \(day_t\) |
\(md^{k}_t\) | The k-th market variable on \(day_t\) |
\(max\{|md^k|\}\) | The maximum value of the k-th market variable |
\(q_{\alpha }\) | The critical value of Nemenyi test |
K | The number of involved algorithms |
\(N_{stock}\) | The number of stocks |
\(T_1\) | The news window |
\(T_2\) | The market data window |
\(w^{T_1}_{t,m}\) | The weight of \(word_m\) obtained from \(T_1\)-days of news |
r(i) | The return obtained from the trading strategy on trading \(day_i\) |
\(signal_i\in \{0,1\}\) | A dummy variable representing the corresponding strategy signal on trading \(day_i\) |
\(\bar{r_{e}}\) | The average of daily excess returns during the simulation period |
\(\sigma _{e}\) | Volatility of daily excess returns during the simulation period |
\(r_{f}(i)\) | The risk-free rate of interest on the trading \(day_i\) |
\(r_{i,j}\) | The daily return of stock j on trading \(day_i\) |
\(signal_{i,j}\in \{0,1\}\) | The corresponding strategy signal of stock j on trading \(day_i\) |