Algorithm 1: ST-SOM

Input: $\left\{u_T\right\}$, $siz{e}^{first\_layer\_SOM},l{r}^{first\_layer\_SOM}$,      $siz{e}^{second\_layer\_SOM},l{r}^{second\_layer\_SOM}$   Output: $spatio-temporalstructure$ 1 $\left\{U_T\right\},d\leftarrow Data\_dependent\_element\_encoding\left(\left\{u_T\right\}\right)$ 2 $Initialise\_first\_layer\_SOM\leftarrow$    $Initialise\_SOM(siz{e}^{first\_layer\_SOM},l{r}^{first\_layer\_SOM},d)$; 3 $First\_layer\_SOM\leftarrow SOM\_learning\_memory(initialise\_first\_layer\_SOM,\left\{U_T\right\})$; 4 $\left\{U_T^{first\_layer\_SOM}\right\}\leftarrow BMU\_seq\left(\left\{U_T\right\}\right)$; 5 $\left\{U_{HD}\right\},D\leftarrow VSA\_encoding\left(\left\{U_T^{first\_layer\_SOM}\right\}\right)$; 6 $Initial\_second\_layer\_SOM\leftarrow$    $Initialise\_SOM(siz{e}^{second\_layer\_SOM},l{r}^{second\_layer\_SOM},D)$; 7 $Second\_layer\_SOM\leftarrow$    $iSOM\_learning\_memory(initialise\_second\_layer\_SOM,\left\{U_{HD}\right\})$;