Subset training

FIGURE 18. Subset training of the learning machine. The weight vector is corrected by reflexion of the decision plane at the mi sc lassified pattern. 

The landslide inventory is naturally event-based. It is not possible to use a multi-temporal landshde inventory as the traditional landslide susceptibiUty analysis. Therefore, the inventory of landsUdes triggered by the earthquake was randomly partitioned into two subsets, training dataset and testing dataset. From the application of ANN model, the landslide susceptibility map is constructed. The results of verification show 81.274% of success rate and 79.915% of prediction rate. The verification results are of high values. It shows that the ANN model can be used as a precise tool in the earthquake triggered landslide susceptibility mapping when a sufficient number of data are available. 

Sejnowski and Rosenberg used two different sets of words for training (1) 1024 words taken from phonetic transcriptions of informal continuous speech by children and (2) a subset of the 1000 most commonly used words selected from Miriam Webster s Pocket Dictionary. NETtalk was trained on a DEC VAX 11/780 minicomputer. 

Calculate a new calibration for this new subset of the original training set. 

Two models of practical interest using quantum chemical parameters were developed by Clark et al. [26, 27]. Both studies were based on 1085 molecules and 36 descriptors calculated with the AMI method following structure optimization and electron density calculation. An initial set of descriptors was selected with a multiple linear regression model and further optimized by trial-and-error variation. The second study calculated a standard error of 0.56 for 1085 compounds and it also estimated the reliability of neural network prediction by analysis of the standard deviation error for an ensemble of 11 networks trained on different randomly selected subsets of the initial training set [27]. 

Ribavirin may be considered for bronchiolitis caused by respiratory syncytial virus in a subset of patients (those with underlying pulmonary or cardiac disease or with severe acute infection). Use of the drug requires special equipment (small-particle aerosol generator) and specifically trained personnel for administration via oxygen hood or mist tent. 

A recent new trend called Active Learning substitutes the often assumed static setting of training and test set in which a learning machine is applied by the probably more realistic scenario of a continuous flow of data. The outcome of experiments influences the choice and generation of subsequent data points [155]. Active Learning provides tools that help select the most promising next subset of data to be subjected to experimentation [156]. 

A better validation strategy, in such cases, is to use three sample subsets a training set, an optimization set, and an evaluation set. The optimization set is used to find the best modeling settings, while the actual reliability of fhe final model is esfimafed by way of a real prediction on the third subset, formed by objecfs fhaf have never influenced the model. The three-set validation procedure should always be used in ANN modeling, which presents a very high risk of overfiffing. 

---