Electromagnetic Signal Classification Based on Class Exemplar Selection and Multi-Objective Linear Programming

نویسندگان

چکیده

In the increasingly complex electromagnetic environment, a variety of new signal types are appearing; however, existing classification (ESC) models cannot handle types. this context, emergence class-incremental learning aims to incrementally update model as categories emerge. paper, an framework based on class exemplar selection and multi-objective linear programming classifier (CES-MOLPC) is proposed in order continuously learn classes incremental manner. Specifically, our approach involves adaptive exemplars considering normalized mutual information classifier. The former used maintain capability for previous by selecting key samples, while latter allow adapt quickly categories. Meanwhile, weighted loss function cross-entropy distillation presented fine-tune model. We demonstrate effectiveness CES-MOLPC method through extensive experiments public RML2016.04c data set large-scale real-world ACARS set. results comparative that can achieve significant improvements over state-of-the-art methods.

برای دانلود باید عضویت طلایی داشته باشید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

On solving possibilistic multi- objective De Novo linear programming

Multi-objective De Novo linear programming (MODNLP) is problem for designing optimal system by reshaping the feasible set (Fiala [3] ). This paper deals with MODNLP having possibilistic objective functions coefficients. The problem is considered by inserting possibilistic data in the objective functions coefficients. The solution of the problem is defined and established under the using of effi...

متن کامل

Providing a Method for Solving Interval Linear Multi-Objective Problems Based on the Goal Programming Approach

Most research has focused on multi-objective issues in its definitive form, with decision-making coefficients and variables assumed to be objective and constraint functions. In fact, due to inaccurate and ambiguous information, it is difficult to accurately identify the values of the coefficients and variables. Interval arithmetic is appropriate for describing and solving uncertainty and inaccu...

متن کامل

Multi-period and Multi-objective Stock Selection Optimization Model Based on Fuzzy Interval Approach

The optimization of investment portfolios is the most important topic in financial decision making, and many relevant models can be found in the literature.  According to importance of portfolio optimization in this paper, deals with novel solution approaches to solve new developed portfolio optimization model. Contrary to previous work, the uncertainty of future retur...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

ژورنال

عنوان ژورنال: Remote Sensing

سال: 2022

ISSN: ['2315-4632', '2315-4675']

DOI: https://doi.org/10.3390/rs14051177