From convex to nonconvex: A loss function analysis for binary classification
- Authors: Zhao, Lei , Mammadov, Musa , Yearwood, John
- Date: 2010
- Type: Text , Conference paper
- Relation: Paper presented at10th IEEE International Conference on Data Mining Workshops, ICDMW 2010 p. 1281-1288
- Full Text:
- Reviewed:
- Description: Problems of data classification can be studied in the framework of regularization theory as ill-posed problems. In this framework, loss functions play an important role in the application of regularization theory to classification. In this paper, we review some important convex loss functions, including hinge loss, square loss, modified square loss, exponential loss, logistic regression loss, as well as some non-convex loss functions, such as sigmoid loss, ø-loss, ramp loss, normalized sigmoid loss, and the loss function of 2 layer neural network. Based on the analysis of these loss functions, we propose a new differentiable non-convex loss function, called smoothed 0-1 loss function, which is a natural approximation of the 0-1 loss function. To compare the performance of different loss functions, we propose two binary classification algorithms for binary classification, one for convex loss functions, the other for non-convex loss functions. A set of experiments are launched on several binary data sets from the UCI repository. The results show that the proposed smoothed 0-1 loss function is robust, especially for those noisy data sets with many outliers. © 2010 IEEE.
Multi label classification and drug-reaction associations using global optimization techniques
- Authors: Mammadov, Musa , Yearwood, John , Aliyea, Leyla
- Date: 2004
- Type: Text , Conference paper
- Relation: Paper presented at ICOTA6: 6th International Conference on Optimization - Techniques and Applications, Ballarat, Victoria : 9th December, 2004
- Full Text: false
- Reviewed:
- Description: E1
- Description: 2003000890
Predicting trading signals of stock market indices using neural networks
- Authors: Tilakaratne, Chandima , Mammadov, Musa , Morris, Sidney
- Date: 2008
- Type: Text , Conference paper
- Relation: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Auckland 1 December 2008 through 5 December 2008 Vol. 5360 LNAI, p. 522-531
- Full Text: false
- Description: The aim of this paper is to develop new neural network algorithms to predict trading signals: buy, hold and sell, of stock market indices. Most commonly used classification techniques are not suitable to predict trading signals when the distribution of the actual trading signals, among theses three classes, is imbalanced. In this paper, new algorithms were developed based on the structure of feedforward neural networks and a modified Ordinary Least Squares (OLS) error function. An adjustment relating to the contribution from the historical data used for training the networks, and the penalization of incorrectly classified trading signals were accounted for when modifying the OLS function. A global optimization algorithm was employed to train these networks. The algorithms developed in this study were employed to predict the trading signals of day (t+1) of the Australian All Ordinary Index. The algorithms with the modified error functions introduced by this study produced better predictions. © 2008 Springer Berlin Heidelberg.
Using links to aid web classification
- Authors: Xie, Wei , Mammadov, Musa , Yearwood, John
- Date: 2007
- Type: Text , Conference paper
- Relation: Paper presented at 6th IEEE/ACIS International Conference on Computer and Information Science, ICIS 2007, Melbourne, Victoria : 11th-13th July 2007 p. 981-986
- Full Text:
- Description: In this paper, we will present a new approach of using link information to improve the accuracy and efficiency of web classification. However, different from others, we only use the mappings between linked documents and their own class or classes. In this case, we only need to add a few features called linked-class features into the datasets. We apply SVM and BoosTexter for classification. We show that the classification accuracy can be improved based on mixtures of ordinary word features and out-linked-class features. We analyze and discuss the reason of this improvement.
- Description: 2003005438