Learning internal representations by error propagation
Parallel distributed processing: explorations in the microstructure of cognition, vol. 1
Neural networks: a systematic introduction
Neural networks: a systematic introduction
Neural Networks: A Comprehensive Foundation
Neural Networks: A Comprehensive Foundation
| Hi-index | 0.00 |
cDNA microarray is a high throughput technology for gene expression analysis. Differing from conventional molecular approaches, which detect molecular targets on a one-by-one basis, cDNA microarray monitors gene expressions of living organisms on a global scale. However, the signal detected by a microarray assay contains a significant amount of noise. Certain types of noise are introduced by the systematic variations that are hardly avoidable by experimental approaches. Significant biological information can only be recognized after the original or raw data sets of microarray assay have been effectively processed. We report here our progress in establishing a Neural Network Normalization (N3) approach to cDNA microarray data processing. With the strong learning ability of the artificial neural network, the trained N3 algorithm is capable of the detection and suppression of systematic variations during microarray data processing and has plasticity in handling both linear and non-linear microarray data sets. The potential of this system in signal processing for other types of biochips, including nucleic acid and non-nucleic acid-based biochips, is yet to be explored.