Decoding beads in a randomly assembled optical nose

In Illumina's technology, the term bead is synonymous with microsensors used in optical arrays. Unlike orderly arranged microarrays, a randomly assembled array would need to be processed via a so-called decoding step, in order to identify the location of each beadtype. Illumina's O-nose technology is radically different from the electronic nose (E-nose) technologies by several factors, e.g., the number of sensors. In an O-nose application, one can easily obtain 2000 usable sensors. The quantity of sensors, however, does come at a price, i.e., the necessity for a decoding procedure. The decoding step plays a challenging role in the O-nose technology. A novel supervised learning technique of decoding randomly assembled arrays, based on subspace classifier method is proposed.