scholarly journals Multi-Positive and Unlabeled Learning

2017 ◽  
Author(s):  
Yixing Xu ◽  
Chang Xu ◽  
Chao Xu ◽  
Dacheng Tao
Keyword(s):  
Unlabeled Data ◽  
Learning Problem ◽  
Step Method ◽  
Practical Applications ◽  
Pu Learning ◽  
Positive And Unlabeled Learning ◽  
One Step ◽  
Real World Datasets ◽  
Multi Class Classification ◽  
The Given

The positive and unlabeled (PU) learning problem focuses on learning a classifier from positive and unlabeled data. Some methods have been developed to solve the PU learning problem. However, they are often limited in practical applications, since only binary classes are involved and cannot easily be adapted to multi-class data. Here we propose a one-step method that directly enables multi-class model to be trained using the given input multi-class data and that predicts the label based on the model decision. Specifically, we construct different convex loss functions for labeled and unlabeled data to learn a discriminant function F. The theoretical analysis on the generalization error bound shows that it is no worse than k√k times of the fully supervised multi-class classification methods when the size of the data in k classes is of the same order. Finally, our experimental results demonstrate the significance and effectiveness of the proposed algorithm in synthetic and real-world datasets.

scholarly journals Online Positive and Unlabeled Learning

2020 ◽  
Author(s):  
Chuang Zhang ◽  
Chen Gong ◽  
Tengfei Liu ◽  
Xun Lu ◽  
Weiqiang Wang ◽  
...  
Keyword(s):  
Online Learning ◽  
Learning Algorithm ◽  
Unlabeled Data ◽  
Sequential Data ◽  
Learning Method ◽  
Pu Learning ◽  
Gradient Based ◽  
Learning Scenarios ◽  
Positive And Unlabeled Learning ◽  
Real World Datasets

Positive and Unlabeled learning (PU learning) aims to build a binary classifier where only positive and unlabeled data are available for classifier training. However, existing PU learning methods all work on a batch learning mode, which cannot deal with the online learning scenarios with sequential data. Therefore, this paper proposes a novel positive and unlabeled learning algorithm in an online training mode, which trains a classifier solely on the positive and unlabeled data arriving in a sequential order. Specifically, we adopt an unbiased estimate for the loss induced by the arriving positive or unlabeled examples at each time. Then we show that for any coming new single datum, the model can be updated independently and incrementally by gradient based online learning method. Furthermore, we extend our method to tackle the cases when more than one example is received at each time. Theoretically, we show that the proposed online PU learning method achieves low regret even though it receives sequential positive and unlabeled data. Empirically, we conduct intensive experiments on both benchmark and real-world datasets, and the results clearly demonstrate the effectiveness of the proposed method.

scholarly journals Positive and Unlabeled Learning with Label Disambiguation

2019 ◽  
Author(s):  
Chuang Zhang ◽  
Dexin Ren ◽  
Tongliang Liu ◽  
Jian Yang ◽  
Chen Gong
Keyword(s):  
State Of The Art ◽  
Ground Truth ◽  
Training Data ◽  
Learning Approaches ◽  
Generalization Error ◽  
Binary Classifier ◽  
Learning Problem ◽  
Pu Learning ◽  
Positive And Unlabeled Learning ◽  
Real World Datasets

Positive and Unlabeled (PU) learning aims to learn a binary classifier from only positive and unlabeled training data. The state-of-the-art methods usually formulate PU learning as a cost-sensitive learning problem, in which every unlabeled example is simultaneously treated as positive and negative with different class weights. However, the ground-truth label of an unlabeled example should be unique, so the existing models inadvertently introduce the label noise which may lead to the biased classifier and deteriorated performance. To solve this problem, this paper  proposes a novel algorithm dubbed as "Positive and Unlabeled learning with Label Disambiguation'' (PULD). We first regard all the unlabeled examples in PU learning as ambiguously labeled as positive and negative, and then employ the margin-based label disambiguation strategy, which enlarges the margin of classifier response between the most likely label and the less likely one, to find the unique ground-truth label of each unlabeled example. Theoretically, we derive the generalization error bound of the proposed method by analyzing its Rademacher complexity. Experimentally, we conduct intensive experiments on both benchmark and real-world datasets, and the results clearly demonstrate the superiority of the proposed PULD to the existing PU learning approaches.

Facile preparation of MoS2/maleic acid composite as high-performance anode for lithium ion batteries

2020 ◽  
Vol 44 (37) ◽  
pp. 15887-15894
Author(s):  
Jingshi Wang ◽  
Zhigang Shen ◽  
Min Yi
Keyword(s):  
Electrochemical Performance ◽  
Lithium Ion Batteries ◽  
High Performance ◽  
Lithium Ion ◽  
Composite Anode ◽  
Step Method ◽  
Practical Applications ◽  
Excellent Electrochemical Performance ◽  
One Step ◽  
Facile Preparation

We propose a facile one-step method to prepare a MoS2 composite anode with excellent electrochemical performance and potential for practical applications in lithium ion batteries.

scholarly journals Positive and Unlabeled Learning via Loss Decomposition and Centroid Estimation

2018 ◽  
Author(s):  
Hong Shi ◽  
Shaojun Pan ◽  
Jian Yang ◽  
Chen Gong
Keyword(s):  
Learning Algorithm ◽  
State Of The Art ◽  
False Negative ◽  
Risk Minimization ◽  
Label Noise ◽  
Hinge Loss ◽  
Pu Learning ◽  
Centroid Estimation ◽  
Positive And Unlabeled Learning ◽  
Real World Datasets

Positive and Unlabeled learning (PU learning) aims to train a binary classifier based on only positive and unlabeled examples, where the unlabeled examples could be either positive or negative. The state-of-the-art algorithms usually cast PU learning as a cost-sensitive learning problem and impose distinct weights to different training examples via a manual or automatic way. However, such weight adjustment or estimation can be inaccurate and thus often lead to unsatisfactory performance. Therefore, this paper regards all unlabeled examples as negative, which means that some of the original positive data are mistakenly labeled as negative. By doing so, we convert PU learning into the risk minimization problem in the presence of false negative label noise, and propose a novel PU learning algorithm termed ?Loss Decomposition and Centroid Estimation? (LDCE). By decomposing the hinge loss function into two parts, we show that only the second part is influenced by label noise, of which the adverse effect can be reduced by estimating the centroid of negative examples. We intensively validate our approach on synthetic dataset, UCI benchmark datasets and real-world datasets, and the experimental results firmly demonstrate the effectiveness of our approach when compared with other state-of-the-art PU learning methodologies.

scholarly journals A Facile, One-Step Synthesis of Silicon/Silicon Carbide/Carbon Nanotube Nanocomposite as a Cycling-Stable Anode for Lithium Ion Batteries

Nanomaterials ◽  
2019 ◽  
Vol 9 (11) ◽  
pp. 1624 ◽  
Author(s):  
Yi Zhang ◽  
Kai Hu ◽  
Yunlei Zhou ◽  
Yingbin Xia ◽  
Nengfei Yu ◽  
...  
Keyword(s):  
Silicon Carbide ◽  
Carbon Nanotube ◽  
Lithium Ion Batteries ◽  
Reduction Process ◽  
Lithium Ion ◽  
Synthesis Process ◽  
Interfacial Region ◽  
Step Method ◽  
Practical Applications ◽  
One Step

Silicon/carbon nanotube (Si/CNTs) nanocomposite is a promising anode material for lithium ion batteries (LIBs). Challenges related to the tricky synthesis process, as well as the weak interaction between Si and CNTs, hinder practical applications. To address these issues, a facile, one-step method to synthesize Si/CNTs nanocomposite by using silica (SiO2) as a reactant via a magnesium reduction process was developed. In this synthesis, the heat released enables the as-obtained Si to react with CNTs in the interfacial region to form silicon carbide (SiC). By virtue of the unique structure composed of Si nanoparticles strongly anchored to conductive CNTs network with stable Si–C chemical bonding, the Si/SiC/CNT nanocomposite delivers a stable capacity of ~1100 mAh g−1 and a capacity retention of about 83.8% after 200 cycles at a current density of 100 mA g−1. Our studies may provide a convenient strategy for the preparation of the Si/C anode of LIBs.

scholarly journals Covariate Shift Adaptation on Learning from Positive and Unlabeled Data

2019 ◽  
Vol 33 ◽  
pp. 4838-4845
Author(s):  
Tomoya Sakai ◽  
Nobuyuki Shimizu
Keyword(s):  
Binary Classification ◽  
Unlabeled Data ◽  
Classification Rule ◽  
Learning Problem ◽  
Mild Conditions ◽  
Importance Weighting ◽  
Pu Learning ◽  
Real World Applications ◽  
Input Sample ◽  
Dataset Shift

The goal of binary classification is to identify whether an input sample belongs to positive or negative classes. Usually, supervised learning is applied to obtain a classification rule, but in real-world applications, it is conceivable that only positive and unlabeled data are accessible for learning, which is called learning from positive and unlabeled data (PU learning). Furthermore, in practice, data distributions are likely to differ between training and testing due to, for example, time variation and domain shift. The covariate shift is a dataset shift situation, where distributions of covariates (inputs) differ between training and testing, but the input-output relation is the same. In this paper, we address the PU learning problem under the covariate shift. We propose an importanceweighted PU learning method and reveal in which situations the importance-weighting is necessary. Moreover, we derive the convergence rate of the proposed method under mild conditions and experimentally demonstrate its effectiveness.

Synthesis and Electrochemical Properties of NiFe-Se/rGO Nanocomposites

2020 ◽  
Vol 13 ◽  
Author(s):  
Rouwei Yan ◽  
Biao Xu ◽  
K. P. Annamalai ◽  
Tianlu Chen ◽  
Zhiming Nie ◽  
...  
Keyword(s):  
Electron Microscopy ◽  
Graphene Oxide ◽  
Energy Storage ◽  
Synergistic Effects ◽  
Nitrogen Adsorption ◽  
Storage Device ◽  
Energy Storage And Conversion ◽  
Step Method ◽  
Practical Applications ◽  
Device Applications

Background : Renewable energies are in great demand because of the shortage of traditional fossil energy and the associated environmental problems. Ni and Se-based materials are recently studied for energy storage and conversion owing to their reasonable conductivities and enriched redox activities as well as abundance. However, their electrochemical performance is still unsatisfactory for practical applications. Objective: To enhance the capacitance storage of Ni-Se materials via modification of their physiochemical properties with Fe. Methods: A two-step method was carried out to prepare FeNi-Se loaded reduced graphene oxide (FeNi-Se/rGO). In the first step, metal salts and graphene oxide (GO) were mixed under basic condition and autoclaved to obtain hydroxide intermediates. As a second step, selenization process was carried out to acquire FeNi-Se/rGO composites. Results: X-ray diffraction measurements (XRD), nitrogen adsorption at 77K, scanning electron microscopy (SEM) and transmission electron microscopy (TEM) were carried out to study the structures, porosities and the morphologies of the composites. Electrochemical measurements revealed that FeNi-Se/rGO notably enhanced capacitance than the NiSe/G composite. This enhanced performance was mainly attributed to the positive synergistic effects of Fe and Ni in the composites, which not only had influence on the conductivity of the composite but also enhanced redox reactions at different current densities. Conclusion: NiFe-Se/rGO nanocomposites were synthesized in a facile way. The samples were characterized physicochemically and electrochemically. NiFeSe/rGO giving much higher capacitance storage than the NiSe/rGO explained that the nanocomposites could be an electrode material for energy storage device applications.

Effect of Dynamic Crosslinking on Mechanical Properties of PP/EPDM Blends

1993 ◽  
Vol 58 (11) ◽  
pp. 2642-2650 ◽  
Author(s):  
Zdeněk Kruliš ◽  
Ivan Fortelný ◽  
Josef Kovář
Keyword(s):  
Mechanical Properties ◽  
Impact Strength ◽  
Shear Modulus ◽  
High Impact ◽  
Necessary Condition ◽  
Step Method ◽  
Melt Mixing ◽  
One Step ◽  
High Impact Strength ◽  
Dynamic Curing

The effect of dynamic curing of PP/EPDM blends with sulfur and thiuram disulfide systems on their mechanical properties was studied. The results were interpreted using the knowledge of the formation of phase structure in the blends during their melt mixing. It was shown, that a sufficiently slow curing reaction is necessary if a high impact strength is to be obtained. Only in such case, a fine and homogeneous dispersion of elastomer can be formed, which is the necessary condition for high impact strength of the blend. Using an inhibitor of curing in the system and a one-step method of dynamic curing leads to an increase in impact strength of blends. From the comparison of shear modulus and impact strength values, it follows that, at the stiffness, the dynamically cured blends have higher impact strength than the uncured ones.

Bismuth particles imbedded degradable nanohydrogel prepared by one-step method for tumor dual-mode imaging and chemo-photothermal combined therapy

2019 ◽  
Vol 375 ◽  
pp. 122000 ◽  
Author(s):  
Yang Xuan ◽  
Xian-Lin Song ◽  
Xiao-Quan Yang ◽  
Ruo-Yun Zhang ◽  
Zi-Yu Song ◽  
...  
Keyword(s):  
Combined Therapy ◽  
Dual Mode ◽  
Step Method ◽  
One Step

scholarly journals Bio-Based Polyurethane Networks Derived from Liquefied Sawdust

Materials ◽  
2021 ◽  
Vol 14 (11) ◽  
pp. 3138
Author(s):  
Kamila Gosz ◽  
Agnieszka Tercjak ◽  
Adam Olszewski ◽  
Józef Haponiuk ◽  
Łukasz Piszczyk
Keyword(s):  
Biomass Conversion ◽  
Structural Features ◽  
Mechanical Analysis ◽  
Process Efficiency ◽  
Hydroxyl Number ◽  
Step Method ◽  
One Step ◽  
Liquefaction Process ◽  
The Fourier Transform ◽  
Liquefaction Temperature

The utilization of forestry waste resources in the production of polyurethane resins is a promising green alternative to the use of unsustainable resources. Liquefaction of wood-based biomass gives polyols with properties depending on the reagents used. In this article, the liquefaction of forestry wastes, including sawdust, in solvents such as glycerol and polyethylene glycol was investigated. The liquefaction process was carried out at temperatures of 120, 150, and 170 °C. The resulting bio-polyols were analyzed for process efficiency, hydroxyl number, water content, viscosity, and structural features using the Fourier transform infrared spectroscopy (FTIR). The optimum liquefaction temperature was 150 °C and the time of 6 h. Comprehensive analysis of polyol properties shows high biomass conversion and hydroxyl number in the range of 238–815 mg KOH/g. This may indicate that bio-polyols may be used as a potential substitute for petrochemical polyols. During polyurethane synthesis, materials with more than 80 wt% of bio-polyol were obtained. The materials were obtained by a one-step method by hot-pressing for 15 min at 100 °C and a pressure of 5 MPa with an NCO:OH ratio of 1:1 and 1.2:1. Dynamical-mechanical analysis (DMA) showed a high modulus of elasticity in the range of 62–839 MPa which depends on the reaction conditions.

Sign in / Sign up

Export Citation Format

Share Document