Image Classiication Using Adaptative-learning Techniques. Automatic Estimation of the Lvq-1 Parameters

| Nearest Neighbor rules are widely used non-parametric classiiers in Pattern Recognition. The main drawbacks of these rules are related to the computational eeort required. In that sense, some techniques have been proposed to select a reduced and representative reference set from the original training set. Adaptative learning techniques may be used successfully to reduce the reference set size. The main drawback of these techniques is the accurate selection of the parameters involved that is required. In this paper, we propose two algorithms for helping to nd accurate values for the parameters involved in LVQ-1 learning, and we point out that this methodology may be extended to diierent adaptative learning techniques.