Double Parallel Extreme Learning Machine

Data driven soft sensor development for complex chemical processes using extreme learning machine

“…However, there are no connections between the input layer nodes and the output layer nodes in ELM. Fortunately, a double parallel structure mentioned in previous work (He et al, 2015a,b;He and Huang, 2005;Yao et al, 2011) can be adopted to solve this problem. The double parallel structure can enable the output layer nodes to not only receive the information from the hidden layer nodes but also receive the direct information from the input nodes, which can enhance the performance of networks (He et al, 2015a,b).…”

Positive and negative correlation input attributes oriented subnets based double parallel extreme learning machine (PNIAOS-DPELM) and its application to monitoring chemical processes in steady state

“…In the paper of Huang et al, an extreme learning machine with kernel (ELMK) which uses unknown kernel mappings instead of the known hidden layer mappings is proposed to avoid the problem of selecting the hidden layer nodes number [16]. In the paper of Yao et al, a double parallel ELM (DP-ELM) was proposed for improving the performance of ELM [41]. The double parallel structure can enable the output nodes to not only receive the information from the hidden layer nodes but also receive the direct information from the input layer nodes [42].…”