Joint Label Inference and Discriminant Embedding
ABSTARCT :
Graph-based learning in semisupervised models provides an effective tool for modeling big data sets in high-dimensional spaces. It has been useful for propagating a small set of initial labels to a large set of unlabeled data. Thus, it meets the requirements of many emerging applications. However, in real-world applications, the scarcity of labeled data can negatively affect the performance of the semisupervised method. In this article, we present a new framework for semisupervised learning called joint label inference and discriminant embedding for soft label inference and linear feature extraction. The proposed criterion and its associated optimization algorithm take advantage of both labeled and unlabeled data samples in order to estimate the discriminant transformation. This type of criterion should allow learning more discriminant semisupervised models. Nine public image data sets are used in the experiments and method comparisons. These experimental results show that the performance of the proposed method is superior to that of many advanced semisupervised graph-based algorithms.
EXISTING SYSTEM :
? A novel constrained affinity matrix construction method is introduced for initial graph construction, which has the powerful ability to excavate the proficiency and complicated structure in the data.
? In this paper, a new constrained graph-based semi-supervised algorithm called constrained label propagation with particle competition and cooperation (CLPPCC) is presented for HSI classification.
? Several experiments were performed to evaluate the learning efficiency of our proposal and demonstrate its superiority when compared to alternatives in this section.
? In this paper, we considered a graph-based semi-supervised problem where the usage of unlabeled samples might deteriorate the model performance in HSI classification.
DISADVANTAGE :
? We consider the problem of inferring node labels in a partially labeled graph where each node in the graph has multiple label types and each label type has a large number of possible labels.
? The concrete problem considered in this paper is that of inferring multiple fields such as the hometowns, current cities, and employers of users of a social network, where users often only partially fill in their profile, if at all.
? We show that this inference problem cannot be split up into separate problems, one for each label type, without loss of accuracy.
? The first uses alternating maximization on a relaxation of the objective, such that each iteration solves a convex problem.
PROPOSED SYSTEM :
• A new active multi-view multi-learner framework based on genetic algorithm (GA-MVML AL) was proposed by Jamshidpour et al., which used the unique high dimensionality of hyperspectral data to construct multi-view algorithms and obtain a more accurate data distribution by multi-learner.
• Therefore, the proposed algorithm has the power to enhance the effectiveness of label propagation, which benefits from the advantages of the constrained graph construction approach.
• With the purpose of estimating the performance of our proposed CLPPCC algorithm, several related algorithms were used for comparative experiments such as TSVM, LGC, and LPA.
ADVANTAGE :
? Although there has been some work on understanding how to combine and weigh different edge types for best prediction performance (Macskassy, 2007), the edge types (analogous to our reason for an edge) were given up front; we recover them automatically.
? We observe similar performance of both methods for hometown and current city, but increasing improvements for high school, college, and employer.
? Social network cluster sizes must display wide variation (e.g., the cluster for users living in New York city versus those in small rural areas), and it is unclear if current network community detection algorithms offer good performance across the entire range of community sizes (Leskovec et al., 2010).
|