31 results about "Class membership" patented technology

Rights management information is used at least in part in a matching, narrowcasting, classifying and/or selecting process. A matching and classification utility system comprising a kind of Commerce Utility System is used to perform the matching, narrowcasting, classifying and/or selecting. The matching and classification utility system may match, narrowcast, classify and/or select people and/or things, non-limiting examples of which include software objects. The Matching and Classification Utility system may use any pre-existing classification schemes, including at least some rights management information and/or other qualitative and/or parameter data indicating and/or defining classes, classification systems, class hierarchies, category schemes, class assignments, category assignments, and/or class membership. The Matching and Classification Utility may also use at least some rights management information together with any artificial intelligence, expert system, statistical, computational, manual, or any other means to define new classes, class hierarchies, classification systems, category schemes, and/or assign persons, things, and/or groups of persons and/or things to at least one class.

A method trains a multi-class classifier by iteratively performing the following steps until a termination condition is reached. The probabilities of class membership for unlabeled data obtained from an active pool of unlabeled data are estimated. A difference between a largest probability and a second largest probability is determined. The unlabeled data with the lowest difference is selected, labeled and then added to a training data set for training the classifier.

Embodiments of a computer-implemented method of calculating estimates of a joint posterior probability of class membership given combinations of attribute values of a pair of attributes are disclosed. The calculating is performed on the basis of data representing a training set of a plurality of instances defined by attribute values for a plurality of attributes together with a class membership outcome. Embodiments of the method comprise calculating first and second estimates of a posterior probability of class membership given attribute values of a first and second attribute of the pair, respectively, and binning the first and second estimates into a respective plurality of first and second probability range bins. Instances of the training set are mapped to combinations of one of each of the first and second pluralities of probability range bins, and on the basis of the mapping, calculating estimates of a joint posterior probability of class membership.

For changing a segmentation of an audio piece after a segment class assignment, at first a short segment is selected, which has a length shorter than a predetermined minimum length. This short segment is preferably merged with the corresponding successor segment or predecessor segment using information on a segment class membership of the short segment itself, but also the successor segment or the predecessor segment, in order to obtain a changed segmentation of the audio signal. With this, a not over-segmented segment representation of the audio signal is obtained, which further includes all audio information, i.e. is not a representation of the audio piece with holes.

The invention, which belongs to the remote sensing data ground object classification field, particularly relates to a precise LINDAR data ground object classification method based on adaptive characteristic weight synthesis. According to the invention, full feature information extraction is carried out on an experiment image, elevation, spectrum, intensity, and texture feature subsets are constructed based on the physical significance of the feature and a difference including ground object information; importance differences of different feature subsets during the ground object classification process are analyzed under a random forest frame, importance measures of all feature subsets are calculated, and class memberships of pixels to all kinds of ground objects are obtained; with comprehensive utilization of the importance measures of the feature subsets as well as an evidence-conflict-calculation-based weight coefficient, synthesis of multiple evidence sources formed by all feature subsets is carried out; and according to the synthesis result, precise ground object classification is realized by using a voting-based decision making rule and a preliminary classification result is optimized by employing an effective space limitation strategy.

Disclosed herein are classifier models, computer implemented systems, machine learning systems and methods thereof for classifying asymptomatic patients into a risk category for having or developing cancer and/or classifying a patient with an increased risk of having or developing cancer into an organ system-based malignancy class membership and/or into a specific cancer class membership.

The invention relates to an enhanced relationship classifier based on representative samples. The method mainly comprises two steps: first, selecting representative samples to form a new training sample set Xnew according to the clustering membership of samples; and then, constructing a fuzzy relationship matrix R with a Phi composite operator about the clustering membership and the class membership of Xnew. The enhanced relationship classifier is mainly characterized in that (1) the matrix R can reveal the inherent logic relationship between a cluster and a class; (2) the computation complexity of the matrix R decreases from O(NLc) to O(MLc), wherein L is the number of classes, c is the number of clusters, N is the number of samples of the original dataset X, M is the number of samples of Xnew, and N is greater than M; and (3) when sufficient judgment information cannot be found in certain areas in the sample space, the classifier rejects to make strategies for test samples falling into the areas, so as to guarantee the confidence level of classification results.

An apparatus, system, and method are disclosed for determining a partial class membership of a data record in a class. The apparatus includes a record set acquisition module that receives a set of reference records having the same independent variables and belonging to a known class within a group of classes. An unknown-class record receiving module receives an unknown-class record having same independent variables as reference records. A class identification module creates a class vector for each reference record identifying whether the record is in a class. A weighting module calculates a set of unknown-class record weights for the unknown-class record. A classification module determines a partial class membership for the unknown-class record for each class in the group of classes using the set of unknown-class record weights. Each partial class membership identifies a probability that the unknown-class record belongs to a corresponding class in the group of classes.

The invention discloses a multi-pose face recognition method based on collaborative fuzzy mean discriminant analysis, which comprises the steps of acquiring a multi-pose face image training sample set comprising a plurality of different classes, performing normalization on each training sample and a sample to be recognized, and performing dimension reduction by using PCA; calculating the class membership degree of each training sample by using a collaboration representation coefficient of the training samples; calculating a fuzzy class mean; calculating the fuzzy intra-class divergence and the fuzzy inter-class divergence of the training samples; solving a projection matrix through maximizing a ratio of the fuzzy inter-class divergence and the fuzzy intra-class divergence of the training samples, and extracting features of the training samples and the sample to be recognized by using the projection matrix; and judging and determining a class label of the sample to be recognized according to a nearest neighbor classifier. According to the invention, class information of the samples are sufficiently utilized, the similarity of the same class of samples and the difference of different classes of samples are considered, and the robustness for noise and wild points is enhanced through introducing membership degree information when a sample has various changes in illumination, pose and expression.

The invention, which belongs to the crossing field of data mining and remote sensing image processing, discloses a multi-spectral remote sensing image terrain classification method. A new membership degree calculation formula is designed and a weighted index is reduced appropriately; a marked sample set for some data is obtained based on the expert knowledge and the centroid is initialized by using the marked sample set; restraining is carried out by using marked samples during the iterative calculation of the centroid and the effect of the marked samples is maximized by using a marker trust factor; and a membership interval is constructed by a fuzzy distance measure, an adaptive factor is constructed by using an intra-class mean square error, and the interval length of the membership function is dynamically adjusted to explore an equivalent I-class membership, and the membership is normalized. With the method provided by the invention, a good classification result is obtained; the robustness and well-posedness to large-scale remote sensing image data are high; and compared with the existing fuzzy C-means method, the provided method has the improved precision.

The invention discloses semi-supervised classification method of modified clustering assumption combined with pairwise constraints. The method relates to a semi-supervised learning algorithm, and includes the steps of: initializing class membership degrees of unlabeled samples through an FCM method; selecting appropriate parameters of lambda1 and lambda2, and calculating initialized alpha, a membership degree function of v(x) and a new objective function of M according to formulas; judging whether an iteration termination condition is reached; if yes, returning the membership degree function v(x), and obtaining a classification decision function of f(x) according to the alpha; and otherwise, re-calculating the initialized alpha, the membership degree function of v(x) and the new objectivefunction of M, and carrying out judgement. According to the method, exploration of modified clustering assumption on the unlabeled samples and utilization of the pairwise constraints on supervisory information are combined to jointly form more completed empirical risk terms, thus knowledge contained by the supervisory information is further mined, and the purpose of algorithm performance improvement is achieved. The method has higher validity and correctness.