Abstract
This chapter proposes a novel method of incremental interference-free neural network training (IIFNNT) for medical datasets, which takes into consideration the interference each attribute has on the others. A specially designed network is used to determine if two attributes interfere with each other, after which the attributes are partitioned using some partitioning algorithms. These algorithms make sure that attributes beneficial to each other are trained in the same batch, thus sharing the same subnetwork while interfering attributes are separated to reduce interference. There are several incremental neural networks available in literature (Guan & Li, 2001; Su, Guan & Yeo, 2001). The architecture of IIFNNT employed some incremental algorithm: the ILIA1 and ILIA2 (incremental learning with respect to new incoming attributes) (Guan & Li, 2001).
Key Terms in this Chapter
Interference: Interference or cross-talking arises during the process of deciding the output value for a particular output attribute, the influence from two or more input attributes are conflicting.
Training: Refers to the process of adjusting or updating neural network weights to deliver the desired output function.
Artificial Intelligence: A branch of computer science that studies human intelligence and aims at embedding intelligent behavior, learning, and adaptation capabilities into machines.
Overfitting: Refers to fitting a model (e.g., neural network) with too many samples or parameters.
Validation: The process of checking whether overfitting (overlearning) is reached during neural network training.
Neural Network: A network of interconnecting neurons working together to produce some output function. The working of a neural network relies on the cooperation of the individual neurons within the network.
Discrimination Ability: The ability of that particular attribute in classification performance when all the other attributes are absent from the neural network training.